the effect of stem surface treatment and substrate
TRANSCRIPT
Western University Western University
Scholarship@Western Scholarship@Western
Electronic Thesis and Dissertation Repository
8-14-2013 12:00 AM
The Effect of Stem Surface Treatment and Substrate Material on The Effect of Stem Surface Treatment and Substrate Material on
Joint Replacement Stability: An In-Vitro Investigation into the Joint Replacement Stability: An In-Vitro Investigation into the
Stem-Cement Interface Mechanics under Various Loading Modes Stem-Cement Interface Mechanics under Various Loading Modes
Yara K. Hosein, The Unviersity of Western Ontario
Supervisor: Dr. Cynthia E. Dunning, The University of Western Ontario
A thesis submitted in partial fulfillment of the requirements for the Doctor of Philosophy degree
in Biomedical Engineering
© Yara K. Hosein 2013
Follow this and additional works at: https://ir.lib.uwo.ca/etd
Part of the Biomechanics and Biotransport Commons
Recommended Citation Recommended Citation Hosein, Yara K., "The Effect of Stem Surface Treatment and Substrate Material on Joint Replacement Stability: An In-Vitro Investigation into the Stem-Cement Interface Mechanics under Various Loading Modes" (2013). Electronic Thesis and Dissertation Repository. 1479. https://ir.lib.uwo.ca/etd/1479
This Dissertation/Thesis is brought to you for free and open access by Scholarship@Western. It has been accepted for inclusion in Electronic Thesis and Dissertation Repository by an authorized administrator of Scholarship@Western. For more information, please contact [email protected].
THE EFFECT OF STEM SURFACE TREATMENT AND SUBSTRATE MATERIAL ON JOINT
REPLACEMENT STABILITY: AN IN-VITRO INVESTIGATION INTO THE STEM-CEMENT INTERFACE
MECHANICS UNDER VARIOUS LOADING MODES
(Thesis format: Integrated Article)
by
Yara Kareen Hosein
Graduate Program in Biomedical Engineering
A thesis submitted in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
The School of Graduate and Postdoctoral Studies
The University of Western Ontario
London, Ontario, Canada
© Yara K. Hosein 2013
ii
ABSTRACT
Mechanical loosening is a common mode of joint replacement failure. For cemented
implants, loosening at the implant-cement interface may be affected by stem surface design.
Altering the surface topography facilitates the infiltration of bone cement onto the stem,
creating a mechanical interlock, improving interface stability. However, few in-vitro studies
have investigated this. Therefore, the purpose of this thesis was to investigate the effect of
stem surface treatments and substrate materials on stem-cement interface stability in-vitro.
Four separate studies were performed to assess the stability of various stem surface
treatments, with two substrate materials, under three loading modes. Titanium and cobalt
chrome implant stems were custom machined and treated with one of four surfaces: smooth,
sintered beads, plasma spray, and circumferential grooves. Sintered bead and plasma
sprayed stems were tested in independent torsion, compression and bending; circumferential
groove designs were compared in torsion and then compression. All stems were potted in
aluminum tubes using PMMA, and loaded cyclically using a materials testing machine. A
custom optical tracking system (resolution under 5 μm) was validated for use, and
subsequently employed to measure stem-cement interface motion during loading. Overall,
results showed surface treatments improved stability, but this was affected by substrate
material. Across all loading modes, beaded treatments applied to titanium stems, and plasma
spray treatments applied to cobalt chrome stems, improved interface stability and strength
when large surface treatment areas were employed. Additionally, the machining of
circumferential grooves onto the stem surface improved interface strength in compression,
with no influence in torsion.
A final study was performed using μ-CT imaging to observe stem and cement motion
under bending loads. A custom-built loading device applied static loads to smooth titanium
stems, while acquiring CT images of the stem-cement interface. Interface motion was
quantified by comparing scans before and after the stem underwent cyclic loading. Results
indicated the stem and the surrounding cement had displaced following loading, yet the stems
remained relatively stable.
iii
These studies offer valuable information regarding the effect of stem surface
treatments on stem-cement interface mechanics under various loading modes and will be
used in the development of future implant systems.
Keywords: joint replacement system; implant design; stem loosening; stem surface
treatment; stem material; bone cement, stem-cement interface; implant stability
iv
CO-AUTHORSHIP STATEMENT
The in-vitro experiments performed in this thesis were multi-discplinary in nature, with the
collaboration of various researchers. The individual contributions are listed below:
Chapter 1: Yara K. Hosein—wrote manuscript; Cynthia E. Dunning—reviewed and
revised manuscript.
Chapter 2: Yara K. Hosein—study design, data collection, data analysis, wrote
manuscript; Cynthia E. Dunning—study design, reviewed and revised
manuscript.
Chapter 3: Yara K. Hosein—study design, data collection, data analysis, wrote
manuscript; Graham J.W. King—study design, reviewed and revised
manuscript; Cynthia E. Dunning—study design, reviewed and revised
manuscript.
Chapter 4: Yara K. Hosein—study design, data collection, data analysis, wrote
manuscript; Graham J.W. King—study design, reviewed and revised
manuscript; Cynthia E. Dunning—study design, reviewed and revised
manuscript.
Chapter 5: Yara K. Hosein—study design, data collection, data analysis, wrote
manuscript; Graham J.W. King—study design, reviewed and revised
manuscript; Cynthia E. Dunning—study design, reviewed and revised
manuscript.
Chapter 6: Yara K. Hosein—study design, data collection, data analysis, wrote
manuscript; Meghan P. Clynick—data collection; Cynthia E. Dunning—study
design, reviewed and revised manuscript.
Chapter 7: Yara K. Hosein—study design, data collection, data analysis, wrote
manuscript; Hristo Nikolov—data collection; Matthew Teeter—study design;
David Holdsworth—study design; Cynthia E. Dunning—study design,
reviewed and revised manuscript.
v
Chapter 8: Yara K. Hosein—wrote manuscript; Cynthia E. Dunning—reviewed and
revised manuscript.
vi
ACKNOWLEDGMENTS
First and foremost, my sincerest gratitude goes to Dr. Cynthia Dunning, who
acknowledged my research potential five years ago, and guided me through the most
formative years of my research. Dr. Dunning has always been a dedicated, patient and caring
supervisor. Her encouraging leadership style, along with her positive and motivational
attitude, has made my graduate studies that much more enjoyable. Dr. Dunning—I thank
you for the wonderful opportunity to have worked with you. You have instilled in me the
qualities and attributes needed to move forward in my research career. I could not have
found a better mentor.
Secondly, I would like to acknowledge my advisory committee members. In
particular, I would like to thank Dr. Graham King, who provided his wealth of clinical
knowledge and research expertise throughout the duration of this thesis. I am extremely
grateful for the time he took from his busy schedule to guide me through my study designs
and manuscript preparations. I would also like thank Dr. Jim Johnson for his overall support
and research advice throughout my graduate studies. I was very fortunate to have worked
with the both of you.
To my amazing lab mates at the Jack McBain Biomechanical Testing Lab, thank you
for your support and assistance over the years. In particular: Dr. Stew McLachlin and Dr.
Tim Burkhart for their abundance of research experience; Jake Reeves for his Solidworks
help and cementing support; Mark Neuert for his engineering wisdom; Sayward Fetterly for
her countless hours of cementing assistance; Meghan Clynick for her assistance with testing
and data collection; and all the past lab members I have had the pleasure to work with.
Thank you for making grad school memorable and enjoyable.
I have also had the opportunity to collaborate with numerous people over the duration
of my graduate studies. I would especially like to thank Dr. David Holdsworth, Dr. Matthew
Teeter and Hristo Nikolov for their contribution to the imaging study within this thesis. In
addition, I would like to acknowledge the staff at University Machine Services, specifically
Clayton Cook and Chris Vandelaar, for lending their expertise.
vii
I must thank the funding sources that supported me through the duration of my PhD
work: The Joint Motion Program (JuMP)—A CIHR training program in Musculoskeletal
Health Research and Leadership and The University of Western Ontario. Thank you for
contributing to my research.
Finally I would like to say a huge thank you to my family. To my parents—thank
you for giving me the opportunity to relocate to Canada to pursue my academic goals, your
infinite love and encouragement will never be forgotten. To Charis and Yanis— thank you
for your cheering me on through my years of never-ending study, your support meant the
world to me.
viii
DEDICATION
To Azard and Yasmin Hosein,
You have supported me endlessly through all of my dreams and aspirations.
ix
TABLE OF CONTENTS
Abstract ............................................................................................................................... ii
Co-Authorship Statement................................................................................................... iv
Dedication ........................................................................................................................ viii
Table of Contents ............................................................................................................... ix
List of Tables ................................................................................................................... xiii
List of Figures .................................................................................................................. xiv
List of Appendices ......................................................................................................... xviii
List of Abbreviations, Symbols and Nomenclature ......................................................... xix
CHAPTER 1: Introduction ................................................................................................. 1
1.1 Joint Replacement Systems..................................................................................... 1
1.1.1 Implantation Technique .............................................................................. 3
1.1.2 Implant Loading .......................................................................................... 7
1.2 Stem Loosening .................................................................................................... 11
1.2.1 Stem Design .............................................................................................. 12
1.2.2 Stem Substrate Material ............................................................................ 18
1.3 Implant Stability.................................................................................................... 21
1.3.1 Implant Stability Measurements in-vitro .................................................. 22
1.4 Optical Systems .................................................................................................... 23
1.4.1 Camera Systems ........................................................................................ 24
1.4.2 Illumination ............................................................................................... 25
1.5 Computed Tomography (CT) Imaging ................................................................. 28
1.6 Study Rationale ..................................................................................................... 29
1.7 Specific Objectives and Hypotheses ..................................................................... 32
1.8 Thesis Overview ................................................................................................... 33
x
1.9 References ............................................................................................................. 35
CHAPTER 2: Development And Validation Of A Custom Optical Tracking System To
Quantify Stem-Cement Interface Micromotion In Stemmed Implant Components .... 48
2.1 Introduction ........................................................................................................... 48
2.2 Materials and Methods .......................................................................................... 49
2.3 Results ................................................................................................................... 55
2.4 Discussion ............................................................................................................. 58
2.5 Conclusion ............................................................................................................ 61
2.6 References ............................................................................................................. 62
CHAPTER 3: The Effect Of Stem Surface Treatment On The Torsional Stability of
Titanium and Cobalt Chrome Cemented Joint Replacement Systems......................... 64
3.1 Introduction ........................................................................................................... 64
3.2 Materials and Methods .......................................................................................... 66
3.3 Results ................................................................................................................... 68
3.4 Discussion ............................................................................................................. 75
3.5 Conclusion ............................................................................................................ 79
3.6 References ............................................................................................................. 80
CHAPTER 4: The Effect Of Stem Surface Treatment And Material On Pistoning Of
Ulnar Components In Linked Cemented Elbow Prostheses ........................................ 84
4.1 Introduction ........................................................................................................... 84
4.2 Materials And Methods......................................................................................... 86
4.3 Results ................................................................................................................... 90
4.4 Discussion ............................................................................................................. 95
4.5 Conclusion .......................................................................................................... 101
4.6 References ........................................................................................................... 102
CHAPTER 5: The Effect Of Stem Circumferential Grooves On The Stability At The
Implant-Cement Interface .......................................................................................... 104
5.1 Introduction ......................................................................................................... 104
xi
5.2 Materials And Methods....................................................................................... 106
5.3 Results ................................................................................................................. 107
5.4 Discussion ........................................................................................................... 109
5.5 Conclusion .......................................................................................................... 117
5.6 References ........................................................................................................... 118
CHAPTER 6: The Effect Of Stem Surface Treatment And Material On The Stability Of
Joint Replacement Systems Subjected To Bending Loads ........................................ 121
6.1 Introduction ......................................................................................................... 121
6.2 Material And Methods ........................................................................................ 123
6.3 Results ................................................................................................................. 125
6.4 Discussion ........................................................................................................... 129
6.5 Conclusion .......................................................................................................... 133
6.6 References ........................................................................................................... 134
CHAPTER 7: The Use of Micro-Computed Tomography (µ-CT) Imaging to Visualize
and Quantify Motion at the Stem-Cement Interface with an Applied Bending Moment
.................................................................................................................................... 136
7.1 Introduction ......................................................................................................... 136
7.2 Materials and Methods ........................................................................................ 138
7.2.1 Testing Specimens .................................................................................. 138
7.2.2 Custom-built Loading Device ................................................................. 138
7.2.3 Loading and Imaging Protocol................................................................ 139
7.2.4 Data Analysis .......................................................................................... 142
7.3 Results ................................................................................................................. 144
7.4 Discussion ........................................................................................................... 149
7.5 Conclusion .......................................................................................................... 153
7.6 References ........................................................................................................... 154
CHAPTER 8: General Discussion and Conclusions ...................................................... 156
xii
8.1 Summary ............................................................................................................. 156
8.2 Strengths and Limitations ................................................................................... 160
8.3 Future Directions ................................................................................................ 162
8.4 Significance......................................................................................................... 164
8.5 References ........................................................................................................... 166
Appendices ...................................................................................................................... 167
Curriculum Vitae ............................................................................................................ 241
xiii
LIST OF TABLES
Table 2.1: Intra-class correlation coefficients (ICC’s) and Standard Errors of Measurement
(SEM) for the Vertical and Horizontal Displacement Measurements of the Optical System. 57
Table 7.1: Pre-dynamic Cemented Stem and Bead Motion ................................................. 146
Table 7.2: Post-dynamic Stem and Bead Motion ................................................................ 147
Table 7.3: Offset Stem and Bead Motion ............................................................................ 148
Table F.1: Vertical and Horizontal Pixel Lengths………………………………………... 185
Table I.1: Tabulated Data for Titanium Surface Treated Implant Stems in Torsion ……...208
Table I.2: Tabulated Data for Cobalt Chrome Surface Treated Implant Stems in Torsion..209
Table I.3: Tabulated Data for Titanium Surface Treated Implant Stems under Compression
............................................................................................................................................... 210
Table I.4: Tabulated Data for Cobalt Chrome Surface Treated Implant Stems under
Compression ......................................................................................................................... 211
Table I.5: Tabulated Data for Circumferential Grooved Implant Stems under Compression
............................................................................................................................................... 212
Table I.6: Tabulated Data for Circumferential Grooved Implant Stems in Torsion ............ 213
Table I.7: Tabulated Data for Titanium Surface Treated Implant Stems in Bending .......... 214
Table I.8: Tabulated Data for Cobalt Chrome Surface Treated Implant Stems in Bending.215
Table L.1: Bending Modulus of Bone Cement with Various Bead- Cement Ratios……....223
Table O.1: Deviation in Stem and Bead Surfaces between Scan- Re-scan Tests………… 240
xiv
LIST OF FIGURES
Figure 1.1: Examples of Synovial Joint Types in the Upper and Lower Limb ....................... 2
Figure 1.2: Joint Replacement System at the Hip .................................................................... 4
Figure 1.3: Methods of Implant Fixation at the Elbow Joint ................................................... 8
Figure 1.4: Loading at the Replaced Shoulder Joint ................................................................ 9
Figure 1.5: Stem Surface Treatments used in Cemented Elbow Systems ............................. 16
Figure 1.6: Schematic of the Plasma Spray Process .............................................................. 17
Figure 1.7: Schematic of the Sintering Process ..................................................................... 19
Figure 1.8: Schematic of a Typical Camera System .............................................................. 26
Figure 1.9: Schematic of a Conventional Lens versus Telecentric Lens ............................... 27
Figure 2.1: Hardware of the Optical Tracking System .......................................................... 51
Figure 2.2: Calibration Grid used for Pixel to mm Conversion ............................................. 52
Figure 2.3: Optical Tracking Software .................................................................................. 53
Figure 2.4: Experimental Set-up for Validation of Optical Tracking System ....................... 54
Figure 2.5: Bland & Altman Plot Comparing Agreement of Measurement Systems ............ 56
Figure 3.1: Surface Treated Stems used for Torsional Testing ............................................. 67
Figure 3.2: Experimental Set-up for Torsional Testing of Implant Stems ............................ 69
Figure 3.3: Stem Rotation for Titanium and Cobalt Chrome Surface Treated Stems ........... 70
Figure 3.4: Inspection of Stem Surface Treatments Post-Torsion Tests ............................... 72
Figure 3.5: Torque at Failure for Surface Treated Stems ...................................................... 73
xv
Figure 3.6: Normalized Interface Toggle for Surface Treated Stems in Torsion .................. 74
Figure 4.1: Surface Treated Implant Stems used for Compression Testing .......................... 87
Figure 4.2: Experimental Set-up for Compression Testing of Implant Stems ....................... 89
Figure 4.3: Inspection of Stem Surface Treatments Post-Compression Tests ....................... 91
Figure 4.4: Survival Curves for Titanium and Cobalt Chrome Stems ................................... 92
Figure 4.5: Stem Motion for Titanium and Cobalt Chrome Stems under Compression ....... 93
Figure 4.6: Interface Stability Offered by the Surface Treated Stems in Compression ........ 94
Figure 4.7: Global Stem Motion of Surface Treated Stems in Compression ........................ 96
Figure 5.1: Smooth and Circumferential Grooved Surface Designs Tested in Both
Compression and Torsion ..................................................................................................... 108
Figure 5.2: Stem Motion for Smooth and Circumferential Grooved Stems ........................ 110
Figure 5.3: Interface Stability for Smooth and Circumferential Grooved Surfaces in
Compression ......................................................................................................................... 111
Figure 5.4: Interface Stability for Smooth and Circumferential Grooved Surfaces in Torsion
............................................................................................................................................... 112
Figure 5.5: Stem Motion with Increased Number of Cycles ............................................... 113
Figure 6.1: Experimental Set-up used for Bending Tests of Implant Stems ....................... 124
Figure 6.2: Stem Motion of Surface Treated Stems under Bending Loads ......................... 126
Figure 6.3: Proximal and Distal Views of Stems Post- Bending Tests ............................... 127
Figure 6.4: Maximum Interface Toggle for Surface Treated Stems in Bending ................. 128
Figure 6.5: Offset Motion for Surface Treated Stems in Bending ....................................... 130
Figure 7.1: Custom-Built Loading Device used for Application of Bending Loads ........... 140
xvi
Figure 7.2: Schematic of Loading and Imaging Protocol used for Static and Dynamic Tests
............................................................................................................................................... 141
Figure 7.3: Surface Deviation Analysis with Geomagic®
................................................... 143
Figure 7.4: Average Stem Motion during Loading for the Pre-and Post-dynamic Scans ... 145
Figure B.1: Specifications for the Basler Pilot AG piA 2400- 12gc Camera…………….. 172
Figure B.2: Specifications for the Opto Engineering Telecentric Lens ............................... 173
Figure B.3: Specifications for the Advanced Illumination, Axial Diffuse Illumintor ......... 174
Figure C.1: Back Panel of Calibration Program………………………………………….. 175
Figure C.2: Back Panel of the Optical Tracking Program ................................................... 176
Figure D.1: Bland Altman plots for Region 1-9 of Horizontal Displacement……………..177
Figure D.2: Bland Altman plots for Region 1-9 of Vertical Displacement ......................... 178
Figure E.1: Line of Best-Fit for Measured against True Displacements…………………. 181
Figure E.2: Deviation Plot for Optical System Measurements ............................................ 182
Figure F.1: Calibration Grid used for Pixel to mm Conversion………………………….. 184
Figure G.1: Engineering Drawing of Smooth Implant Stem……………………………....188
Figure G.2: Engineering Drawing of 20 mm Length Plasma Spray Implant Stem ............. 189
Figure G.3: Engineering Drawing of 20 mm Length Beaded Implant Stem ....................... 190
Figure G.4: Engineering Drawing of 10 mm Length Plasma Spray Implant Stem ............. 191
Figure G.5: Engineering Drawing of 10 mm length Beaded Implant Stem ........................ 192
Figure G.6: Engineering Drawing of 0.6 mm Grooved Implant Stem ................................ 193
Figure G.7: Engineering Drawing of 1.1 mm Grooved Implant Stem ................................ 194
xvii
Figure G.8: Engineering Drawing of Parts for Potting Jig Assembly ................................. 195
Figure G.9: Engineering Drawing of Delrin® Block used to Centralize Stems for Potting.196
Figure G.10: Engineering Assembly of Jig used for securing Stem in Torsion and
Compression ......................................................................................................................... 197
Figure G.11: Engineering Drawing of Securing Plates used for Bending Tests ................. 198
Figure G.12: Engineering Assembly of Cement Sample Template ..................................... 199
Figure G.13: Engineering Drawing of Parts for the Cement Sample Template .................. 200
Figure G.14: Engineering Assembly of Stem Holder used in µCT-imaging Bend Tests .... 201
Figure G.15: Engineering Drawing of Parts for Stem Holder Assembly used in µCT-imaging
Bend Tests ............................................................................................................................. 202
Figure G.16: Engineering Drawing of Back Plate for Stem Holder Assembly used in µCT-
imaging Bend Tests............................................................................................................... 203
Figure H.1: Potting Jig used for Securing Stems during Cementing……………………....205
Figure L.1: Bead Embedded Cement Samples…………………………………………….221
Figure L.2: Experimental Set-up for Three-point Bending of Cement Samples…………. 222
Figure M.1: Mechanical Parts of Loading Device………………………………………... 228
Figure M.2: Custom-Built Loading Device………………………………………………. 229
Figure N.1: Region of Interest (ROI) used to Create Isosurfaces……………………….....232
Figure N.2: Geomagic® Surface Deviation Analysis………………………………………237
xviii
LIST OF APPENDICES
Appendix A: Thesis Glossary .............................................................................. 168
Appendix B: Specifications for Hardware of Optical System ............................. 172
Appendix C: LabVIEW®
Programs ..................................................................... 175
Appendix D: Bland-Altman Plots for Optical System Validation ....................... 177
Appendix E: Calculation of Optical System Accuracy ....................................... 179
Appendix F: Calculation of Optical System Resolution ...................................... 183
Appendix G: Engineering Drawings and Assemblies .......................................... 188
Appendix H: Cement Preparation and Potting Technique ................................... 204
Appendix I: Tabulated Data ................................................................................ 208
Appendix J: Letter of Permission ....................................................................... 216
Appendix K: Bead Embedment Procedure .......................................................... 218
Appendix L: Mechanical Testing of Beaded Cement Samples ........................... 219
Appendix M: Custom-built Loading Device ........................................................ 226
Appendix N: Surface Deviation Analysis ............................................................ 230
Appendix O: Uncertainty of Stem and Bead Displacements ............................... 238
xix
LIST OF ABBREVIATIONS, SYMBOLS AND NOMENCLATURE
Ø diameter
º degrees
% percent
µ micro
µm micrometer
°C degrees Celsius (unit of temperature)
ANOVA analysis of variance
CCD charged couple device
CT computed tomography
DIC digital image correlation
F applied force
FOV field of view
Fres resultant force
Fx component of force in x direction
Fy component of force in y direction
Fz component of force in z direction
g grams
GSM global stem motion
HA hydroxyapatite
xx
HU Hounsfield unit
Hz unit of frequency
i.e., ‘that is’
ICC intraclass correlation coefficient
ISO International Organization for Standardization
IT interface toggle
mm millimeters
N newton (unit of force)
Nm newton meter (unit of torsional load)
p probability value
pixel pixel element
PMMA polymethymethacrylate
Ra average roughness value
ROI region of interest
Rz mean roughness depth
s second (unit of time)
SD standard deviation
voxel volume element
α alpha (significance level of a test)
1
CHAPTER 1: INTRODUCTION
Overview: Stemmed joint replacement systems play an important role in the
treatment of diseased or damaged joints; however, their success depends on the stability
of the system after implantation, and its ability to withstand implant loosening over time.
As such, the overall goal of this thesis was to investigate the stability at the implant-
cement interface, using mechanical testing and an imaging technique, to determine the
contribution of stem surface designs and substrate material in resisting the onset of
implant loosening. This chapter introduces stemmed components of joint replacement
systems, describes various stem design parameters considered for successful
implantation, overviewing techniques used in this thesis for quantifying implant stability,
and concludes with the study rationale, objectives and hypotheses.
1.1 JOINT REPLACEMENT SYSTEMS
Joint replacement systems (i.e., implant systems) are used in field of orthopaedic
surgery as a treatment option for joints affected by degenerative diseases such as
osteoarthritis, osteoporosis or rheumatoid arthritis, where the articular cartilage of the
joint is destroyed. The damage imposed by these diseases is generally considered as
irreversible. As such, replacement of the damaged joint is required. Similarly, traumatic
injury to joints requires removal of the destroyed anatomy, to be replaced with an
artificial prosthesis that restores the joint structure and function.
Implant systems are specific to the joint being replaced. In particular, stemmed
joint replacement systems are unique to the treatment of synovial joints. These include
joints of the lower (i.e. hip, knee, and ankle) and upper (i.e. shoulder, elbow, and wrist)
extremities (Figure 1.1). Compared to the other joints of the body (i.e., cartilaginous and
2
Figure 1.1: Examples of Synovial Joint Types in the Upper and Lower Limb
The above schematic shows examples of the synovial joints found in the human body; (1)
hinge joint found between the humerus and ulna at the elbow, (2) saddle joint between
the thumb and wrist, (3) plane joint found between the bones of the tarsus in the foot, (4)
ball-and-socket joint located between the femoral head and acetabulum at the hip, (5)
condyloid joint found between the bones at the wrist, and (6) pivot joint found between
the head of the radius and radial notch of the ulna at the elbow. (Joint images modified
from Tortora and Nielson, 2009)
3
fibrous joints), synovial joints provide a large range of motion (Tortora and Nielsen,
2009).
During the joint replacement procedure, the damaged or diseased joint is removed
using established surgical techniques, and the host bone is prepared to accommodate the
artificial joint. This is done by resecting the damaged joint head, and creating a canal
through the remaining bone. The surgeon then inserts the stem of the artificial joint into
host bone canal to allow for implant fixation.
Stemmed replacement systems typically consist of two main parts; the articulating
component and intramedullary stem. The articulating region makes up the bearing
surface, and depending on the degrees-of-freedom in the joint being replaced, may be
classified as; (i) hinge joint, (ii) saddle joint, (iii) plane joint, (iv) ball-and-socket joint,
(v) condyloid joint, and (vi) pivot joint (Figure 1.1) (Tortora and Nielsen, 2009). The
intramedullary stem does not mimic a particular structure of the natural joint, but instead
has a role in implantation of the artificial joint. That is, the stemmed region is inserted
into the reamed host bone canal, and allows for fixation and anchoring of the replacement
system to the healthy bone, which remains once the damaged articulating portion is
removed (Figure 1.2).
The success of these replacement systems to stay implanted over time (i.e., to avoid
loosening) depends on the ability of the stemmed component to keep fixated within the
host bone canal. Therefore, to improve on the longevity of joint replacement systems, the
fixation process of the intramedullary stem needs to be thoroughly explored, with the
purpose of ensuring secure anchorage of the implant to the healthy bone.
1.1.1 IMPLANTATION TECHNIQUE
There are two methods used to create a mechanical connection at the implant-
bone interface; biologic fixation and cemented fixation (Figure 1.3). The choice of
4
Figure 1.2: Joint Replacement System at the Hip
(A) Healthy hip joint, and (B) replaced hip joint. During joint replacement surgery, the
damaged or diseased joint is removed and replaced with the artificial joint. The joint
replacement consists of the articulation region that replaces the structure and function of
the joint. The intramedullary stem is inserted into the canal of the host bone and allows
for implant fixation. (Drawing of hip provided by Dr. Angela Kedgley)
5
fixation method may be dependent on implant design, surgeon preference, and patient
specificity (Bauer and Schils, 1999; Morshed et al., 2007; Ni et al., 2005; Park, 1992).
1.1.1.1 BIOLOGIC FIXATION
Biologic fixation involves the osseointegration, or direct structural and functional
connection of the bone with the implant surface (Brånemark et al., 2001). The success of
osseointegration as a method for implant fixation is dependent on the material properties
and the design of the implant stem (Bauer and Schils, 1999; Skinner, 2006). In
particular, implant stem designs incorporating rough finishes, porous surfaces, and
bioactive coatings have been utilized in biologic fixation (Bauer and Schils, 1999).
Rough or porous surfaces are applied to the implant surface to promote bone
growth and ensure biological anchoring of the healthy bone to the implant (Incavo et al.,
2004; Park, 1992). These surface designs have been explored with regards to optimum
surface roughness and pore size needed to optimize bone ingrowth (Bobyn et al., 1980;
Engh and Bobyn, 1988).
In addition to roughened surfaces, the application of bioactive coatings has also
been used in biologic fixation. Incorporating a bioactive coating such as hydroxyapatite
(HA) facilitates stimulation of bone formation on the stem surface (Geesink et al., 1988),
and its osetoconductivity is dependent on the coating composition, density, thickness and
texture (Bauer and Schils, 1999).
Biologic fixation is typically used in younger patient populations, where the quality
the healthy bone can successfully accommodate ingrowth onto the stem surface.
Additionally, depending on the stage of joint replacement surgery (i.e., initial
replacement versus revised replacement), along with the joint being replaced (i.e., upper
limb versus lower limb), biologic fixation may be preferred.
6
1.1.1.2 CEMENTED FIXATION
Orthopaedic bone cement is another method used for stem fixation. This method
involves the incorporation of bone cement into the reamed bone canal, followed by the
subsequent insertion of the intramedullary stem of the joint replacement system, ensuring
appropriate alignment during implantation. Once inserted, a cement mantle thickness of
approximately 2–3 mm (Banaszkiewicz, 2009a) is generated at the fixation interface,
matching the surface of the implant to the bone canal. Cemented fixation is typically
performed in the older patient population, where compromised bone quality may affect
the success of implantation. In addition, a major advantage of using this type of fixation
is the shorter recovery time associated with its procedure, allowing almost immediate
weight bearing at the replaced joint.
One example of the cement used for fixation is polymethylmethacrylate (PMMA)
bone cement, which is a synthetic polymer that works to secure the implant into the bone
canal. PMMA bone cement is packaged as a polymethylmethacrylate powder and a
monomer methacrylate liquid. When the powder and liquid are mixed together, the
monomer is polymerized in a free radical process, to form the viscous paste that hardens
over time (Navarro et al., 2008; Serbetci and Hasirci, 2004; Webb and Spencer, 2007).
Mixing of bone cement can be done using hand mixing techniques or under vacuum
pressure (Dunne et al., 2004; Geiger et al., 2001; Lewis, 1997). After implantation, the
PMMA bone cement does not adhere the implant to the bone, but instead enters the space
between the bone and implant, providing a mechanical connection between the implant
and the bone (Janssen et al., 2008; Scheerlinck and Casteleyn, 2006). As such, when
introduced into the bone canal, the cement conforms itself to the shape of the canal as
well as the shape of the implant stem, creating a ‘customized’ prosthesis fit (Wroblewski
et al., 2008).
There are four main phases that occur during cement preparation and curing. Phase
1, or the mixing phase, begins immediately after the powder and liquid monomer are
combined. When the mixture becomes fully integrated and sticky in consistency, the
start of Phase 2 begins. When the sticky consistency is lost and the mixture starts to
7
appear doughy, this signals the start of Phase 3, the duration of which is termed the
working time. The final phase occurs when the mixture can no longer be manipulated,
but instead begins to harden, signaling Phase 4 (Kuhn, 2009; Serbetci and Hasirci, 2004).
Typically the stem is inserted during Phase 3, and the exact time point of the working
phase in which the insertion takes place is important to the success of stem implantation
(Iesaka et al., 2003; Park, 1992; Smeds et al., 1997). A longer working time, with
doughy cement consistency, facilitates better positioning and support of the implant
(Smeds et al., 1997); however, a shorter working time, with lower viscosity cement may
be useful to allow for better intrusion of the bone cement into interface spaces (Hansen
and Jensen, 1990).
In addition to the different phases of cement preparation and curing, the physical
properties of the PMMA bone cement during those phases play an important role in
implant fixation (Lewis, 1997; Saha and Pal, 1984). Porosity and viscosity are two such
properties that can affect the static and dynamic loading response of bone cement (Lewis,
1997; Saha and Pal, 1984; Verdonschot and Huiskes, 1995). These properties have been
shown to be affected by the temperature during preparation and handling, as well as
mixing method of the cement just prior to implantation (Dall et al., 2007; Hernigou et al.,
2009; Macaulay et al., 2002; Smeds et al., 1997). As such, it is important to ensure a
controlled environment for the preparation of bone cement during joint replacement
implantation.
1.1.2 IMPLANT LOADING
Mobility of the replaced joint is facilitated by muscles, tendons and ligaments
acting together to both stabilize the joint while still allowing motion. Forces exerted by
the individual muscles and ligaments, along with external loads supported by the limb,
work collectively to produce resultant forces acting at the replaced joint. These resultant
forces generate torsion, tension/compression and bending, which typically occur in a
combined state (Figure 1.4). These loads have been shown to be up to three times body
8
Figure 1.3: Methods of Implant Fixation at the Elbow Joint
Shown above are examples of implant fixation at the elbow joint. Non-cemented fixation
incorporates a porous coating onto the surface of the implant stem (humerus), allowing
ingrowth of the surrounding healthy bone. In comparison, cemented fixation uses
orthopaedic bone cement as a filler to match the surfaces of the implant stem (ulna) to
that of the healthy bone. (Image of replaced elbow adapted from www.orthogate.org.)
9
Figure 1.4: Loading at the Replaced Shoulder Joint
Resultant forces (Fres) occur at the joint due to internal muscle and ligamentous forces,
and external joint loads. These forces generate torsion, compression and bending, which
typically occur in a combined state at the joint. (Image of replaced shoulder adapted from
www.orthogate.org.)
10
weight for many joints (i.e., elbow, hip and knee) during daily activities, and even higher
for strenuous activities (Amis et al., 1980; Bergmann et al., 2004, 1993; Kutzner et al.,
2010).
Torsional loads refer to those that occur around the longitudinal axis of the
implanted stem (i.e., z-axis of Figure 1.4). This type of loading is common at the hip
joint, where studies done by Bergmann et al., have shown that torque about an
instrumented prosthesis ranged between 5 Nm and 17 Nm for routine daily activities
(Bergmann et al., 2001, 1993). Daily activities included walking, sitting and
ascending/descending stairs, with the highest torques occurring during stair ascent
(Bergmann et al., 2001). At joints such as the knee, shoulder and elbow, these torsional
loads are present in lower magnitudes than at the hip (Guerra, 2004; Kutzner et al., 2010;
Westerhoff et al., 2009), but can increase during strenuous joint activity.
In addition to torsional loading, all joints are exposed to axial loads (i.e., tensile and
compressive forces) which act along the length of the implant stem (Figure 1.4). At the
lower limb joints such as the hip and knee, compressive loads are apparent due to the
weight bearing nature of the joint. At the hip, these loads can be up to 2000 N during
normal gait (Bergmann et al., 2001, 1993), and up to 7000 N during stumbling
(Bergmann et al., 2004). At the tibial component of the knee joint, compressive forces of
up to 3000 N have been recorded for daily activities (Kutzner et al., 2010). Upper limb
joints such as the elbow and shoulder experience similar compressive loads, despite being
considered non-weight bearing joints (Amis, 2012; Amis et al., 1980; Bergmann et al.,
2007; Johnson and King, 2005). For the shoulder joint, a force of one times body weight
is generated across the glenohumeral joint during 90º of shoulder abduction, and up to 2.5
times body weight when lifting a weight of approximately 50 N (Bergmann, 1987).
Similarly, during 90º elbow flexion carrying a weight, the compressive force at the
ulnohumeral joint can be up to six times the external load at the hand, and even higher
during elbow extension (Amis, 2012; Amis et al., 1980). It is important to note that in
addition to compressive forces at the replaced ulnohumeral joint, tensile forces may also
be apparent. These tensile forces act along the length of the unlar component of linked
elbow prostheses during elbow hyperflexion. The effect of these forces may also be
11
exaggerated when impingement occurs between bony structures or cement, and the
anterior flange of the implant, further creating distraction forces along the length of the
ulnar component (Cheung and O’Driscoll, 2007).
Bending loads applied to implant occur as a result of forces acting perpendicular to
the longitudinal axis of the implant (i.e., Fx, Fy of Figure 1.4). These forces, although
smaller than those axially, have been reported to be approximately 0.5 times body weight
at the hip and shoulder during routine activities (Bergmann et al., 2007, 2001). At the
elbow and knee, these forces were shown to be lower in magnitude (Guerra, 2004;
Kutzner et al., 2010).
Taking into consideration the various types and magnitudes of loads that occur at
the joint, it is necessary to assess the ability of implanted systems to withstand these
loads, in order to determine the overall success of joint replacement systems.
1.2 STEM LOOSENING
Although cemented replacement systems have been a successful treatment option,
stem loosening is the most common method of implant failure (Australian Orthopaedic
Association, 2010; New Zealand Orthopaedic Association, 2010). Stem loosening occurs
when there is disruption to the mechanical connection at the stem-cement interface, as a
result of continuous loading that occurs at the joint, which is transferred to the fixation
interface. There are a variety of factors that can contribute to stem loosening such as
patient profile, surgical technique, and implant stem design (Barrack, 2000; Harris and
Tarr, 1979; Rodriguez-Gonzalez, 2009). Stem design, in particular, can be explored by
biomechanical analysis.
12
1.2.1 STEM DESIGN
Stem design is important for ensuring adequate implant fixation and mechanical
stability to the implant system (Barrack, 2000; Mohler et al., 1995; Scheerlinck and
Casteleyn, 2006). Since the stem undergoes a mechanical connection with the cement
during fixation, altering its design to improve that connection can play a key role in stem
stability, and resistance to stem loosening.
1.2.1.1 STEM SHAPE AND LENGTH
Stem shape is one factor of stem design that can contribute to the mechanical
stability of the implant system. Over the years, many studies have investigated the role
of stem shape on implant stability, and proposed shape designs that are specific to joint
type and loading (Einsiedel et al., 2008; Evans et al., 1988; Huiskes et al., 1998; Kedgley
et al., 2007; Olofsson et al., 2006; Westphal et al., 2006). Stem cross-sectional shape and
curvature are two examples of stem shape that have been proposed to provide implant
stability under torsional loading (Berzins et al., 1993; Callaghan et al., 1992;
Crowninshield et al., 2006; Hosein et al., 2012; Kedgley et al., 2007). With regards to
cross-sectional shape, it has been explained that stems with longer edges provided the
greatest stability under torsional loading (Kedgley et al., 2007). Likewise, the
longitudinal curvature of stem can work to improve the torsional stability of the implant
(Berzins et al., 1993; Callaghan et al., 1992), however, this may be dependent on the
degree of stem curvature (Hosein et al., 2012).
Stem length is another factor that is important in the design of implants. The length
of the implant stem can influence the load transfer to the cement and bone. A study by
Mann et al., demonstrated that shorter stems contributed to higher cement mantle
stresses, which can result in cement damage and initiation of implant loosening (Mann et
al., 1997). With regards to interface fixation, the length of the stem may also affect
implant stability, where a longer stem increases the length of the fixation interface,
thereby increasing the stem-cement connect, and improving initial implant stability.
13
However, this is often at the expense of decreased stresses to bone, which may ultimately
lead to bone loss around the implant and subsequent loosening through the process
known as stress shielding (Austman et al., 2007).
1.2.1.2 STEM SURFACE MODIFICATION
In addition to the stem’s shape and length, stem surface modifications such as
surface finish and surface treatments can also play a role in stabilization of cemented
implant systems. Stem surface finishes have been investigated in cemented implant
designs with the premise that a roughened surface would contribute frictional resistance
to interface loading (Davies and Harris, 1993; Huiskes et al., 1998; Jamali et al., 2006).
Likewise, implant surface treatments are used with some cemented stem designs (Van der
Lugt and Rozing, 2004) to accommodate infiltration of bone cement onto the textured
stem surface, providing improved fixation and stability (Jeon et al., 2012). Both methods
are aimed at providing mechanical resistance to loading at the stem-cement interface.
1.2.1.2.1 STEM SURFACE FINISH
The role of stem surface finishes in implant stability has been described based on
their surface characteristics. These characteristics include the stem’s average surface
roughness (Ra), and mean roughness depth (Rz). The Ra value is the arithmetic average of
all departures from the center line of the roughness profile, the center line being located
where the area of the roughness profiles above and below are equal (Crowninshield,
1998). The Rz is the mean of the depths (highest peak to lowest valley) of five
consecutive sample lengths within the roughness profile. Ra values are, however, more
commonly used to describe implant surfaces. A review by Verdonschot classified
implant surfaces based on these values; smooth (Ra < 1 µm), matte (Ra < 2 µm), and
rough surfaces (Ra > 2 µm) (Verdonschot, 2005).
14
Although rougher surface finished implants have been shown to increase interface
strength due to their frictional resistance to loading, additional studies have argued that
the micro-roughened surfaces of the implant stem also contributed to abrasion and
fretting at the stem-cement interface (Beksac et al., 2006; Crowninshield et al., 1998;
Hinrichs et al., 2003; Mohler et al., 1995). Interface abrasion would ultimately promote
the onset of loosening, in addition to causing concern for the patient as a result of the
accumulation of cement and/or metal debris at the replaced joint. As such, there has been
mixed reviews regarding the use of surface finished stems in cemented implant systems.
1.2.1.2.2 STEM SURFACE TREATMENT
In comparison to surface finishes, stem surface treatment is the alteration of the
surface topography of the implant stem. Surface treatments can involve the addition of
material (i.e., plasma spray and sintered beads), or removal of material (i.e., machined
surface patterns) from the implant stem surface. Both methods change the surface profile
of the stem, accommodating infiltration of bone cement onto the stem surface.
Surface treatments such as plasma spray and sintered beads involve the addition of
porous metal coatings onto the implant stem (Bundy and Penn, 1987; Glass and
Pierfrancesco, 2011; Ryan et al., 2006). Initially, surface treatments were incorporated
into non-cemented implant designs to allow for bony ingrowth, however, within recent
years, cemented elbow implant designs have incorporated surface treatments to enhance
the mechanical interlock between the stem and cement (Evans et al., 1988; Skytta et al.,
2009; van der Lugt and Rozing, 2004) (Figure 1.5). The type of surface treatment, along
with its fabrication process, can dictate the porosity of the stem’s surface to allow for
infiltration of bone cement.
15
1.2.1.2.2.1 PLASMA SPRAY FABRICATION
Plasma spraying is a surface treatment used to create macro-roughened surface
textures on implant stems. The roughness value (Ra) associated with plasma spray
treatments, however, are greater than those defined by Verdonschot et al., to describe
micro-roughed surfaces (Verdonschot, 2005).
During the plasma spraying process, an electric arc is generated by a high voltage
discharge, and formed between two water-cooled electrodes (see Davis, 2003). The
electric arc heats plasma gas flowing through the electrodes to extreme temperatures, as
high as 22000 ºC, partially ionizing the gas to form plasma. The powder used for coating
the implant surface is introduced into the plasma gas stream using a carrier gas, and the
mixture is accelerated at high speeds onto the stem substrate material (Figure 1.6). The
degree of roughness of the coating can be varied by adjusting the spraying parameters of
the process (Ryan et al., 2006). For application of titanium plasma spray coatings,
controlled atmospheric plasma spray (CAPS) or vacuum plasma spray (VPS) are used.
Titanium is extremely sensitive to oxidization in high temperature environments, and as
such, the CAPS units applies the coating in a positive pressure inert atmosphere, while
the VPS unit applies the coating in controlled low vacuum pressure (Glass and
Pierfrancesco, 2011). The plasma spray coatings prepared with this method result in
irregular pores that facilitate a mechanical connect between the stem surface and bone
cement during implant fixation.
1.2.1.2.2.2 SINTERED BEAD FABRICATION
Another method for creating roughened stem surfaces is the attachment of bead
particles to the stem surface, thereby creating a porous network through which bone
cement can fill the spaces. Sintered bead coatings are applied to implant stem by a
process that involves binding and sintering metal beads onto the stem surface (Davis,
2003; Ryan et al., 2006). A binder is first used to attach the metal beads to the stem
16
Figure 1.5: Stem Surface Treatments used in Cemented Elbow Systems
Three commercially available elbow systems that incorporate stem surface treatments
onto their cemented implant designs. (A) Coonrad/Morrey I Total Elbow (Zimmer) with
sintered beads on a titanium substrate (www.zimmer.com), (B) Discovery Elbow System
(Biomet) with plasma spray on a titanium substrate (www.biomet.com), and (C) Latitude®
EV Total Elbow (Tornier) with plasma spray on a cobalt chrome substrate (www.tornier-
us.com).
17
Figure 1.6: Schematic of the Plasma Spray Process
An electric arc is generated by a high voltage discharge formed between the cathode and
anode within the chamber. This arc heats the plasma gas flowing into the chamber,
partially ionizing the gas to form plasma. The titanium powder used for coating the
implant is incorporated into the plasma gas stream, and the mixture is accelerated at high
speeds onto the stem substrate. (Image of implant: Discovery Elbow System (Biomet);
www.biomet.com.)
18
substrate material. The attached beads are then sintered at high temperatures in a high
vacuum oven, where the binder is vaporized, and the beads are diffused onto the stem
substrate material, as well as with each other. The sintering process generally involves
heating the implant stem to one half times the melting point of the metal alloy, to allow
for diffusion of the beads to the stem surface and one another (Davis, 2003) (Figure 1.7).
Typically, the same stem substrate metal is needed to make the beads that will be sintered
to the stem surface (i.e., cobalt chrome beads on cobalt chrome stem, or titanium beads
on titanium stems). The volume fraction of sintered bead surfaces can be altered by
changing the bead size as well as the number of layers of bead coating on the stem. It can
be controlled by variables such as compacted powder density, sintering temperature and
time, and alloy additions (Ryan et al., 2006). The pore size, morphology, distribution of
the beads, and the inter particle neck size can all have a major impact on the mechanical
properties of the resulting coating (Davis, 2003; Ryan et al., 2006).
1.2.2 STEM SUBSTRATE MATERIAL
Implant stems utilize metal alloys as their base substrate material (Korkusuz and
Korkusuz, 2004; Navarro et al., 2008), with the goal of providing a suitable material
replacement for the resected bone. Commonly used metals include iron-based, cobalt-
based and titanium-based alloys. These alloys are chosen based on their
biocompatibility, strength, wear, and corrosion characteristics. Within recent years,
however, the use of iron-based alloys in stem design has been restricted because of
inferior mechanical and corrosion properties compared to cobalt-based and titanium-
based alloys (Navarro et al., 2008).
Cobalt-based alloys are useful as an orthopaedic implant material due to their high
fatigue strength, and high ultimate tensile strength, which make them suitable for their
longevity and ability to resist fracture (Buechel et al., 2012; Korkusuz and Korkusuz,
2004; Navarro et al., 2008). With regards to its biocompatibility, the addition of
chromium is known to inhibit corrosion (Rodriguez-Gonzalez, 2009), and have also
19
Figure 1.7: Schematic of the Sintering Process
Sintered bead coatings are applied to the stem in a process that involves both binding and
sintering the metal beads to the stem’s surface. A binder is incorporated with the beads
to hold the particles together, while high temperatures are used to diffuse the beads onto
the stem surface, and with one another. Temperatures used for sintering are at least one
half times the melting temperature of the metal alloy. The resultant coating is a porous
structure, through which bone and cement can fill the spaces (Note: Picture at the top is
from an SEM scan at x50 magnification.)
20
shown good resistance to wear (Navarro et al., 2008). In addition, the introduction of
small quantities of molybdenum and tungsten has also been used to harden the cobalt
chrome alloys (Rodriguez-Gonzalez, 2009).
Titanium-based alloys were introduced into stem design because of their low
density, moderate elastic modulus, and good corrosion resistance, which made them
biocompatible to the surrounding environment (Hallab et al., 2004; Navarro et al., 2008).
The corrosion resistant nature of titanium is due to an oxide layer that is formed on the
surface, which acts as a protective barrier to corrosion of the material substrate beneath it.
Although titanium is known to exhibit poor shear strength and wear resistant properties,
the addition of aluminum and vanadium in orthopaedic applications improves the
mechanical properties of the metal alloy (Navarro et al., 2008; Rodriguez-Gonzalez,
2009).
However, there has been much debate between the choices of titanium or cobalt
chrome stem material. Within recent years, titanium has been a controversial material for
use with femoral cemented stem designs, since it has been reported to experience
increased rates of loosening, as well as high sensitivity to corrosion (Boyer et al., 2009;
Maurer et al., 2001; Thomas et al., 2004; Willert et al., 1996). It is thought that
micromotion of the stem within the cement leads to abrasion of the titanium protective
oxidized layer, resulting in exposure and gradual corrosion of the base titanium material
over time (Hallam et al., 2004). In contrast, non-cemented titanium femoral stems are
believed to be less prone to corrosion because oxygen is readily available from the
surrounding bone, allowing for a stable oxide-layer to re-accumulate on the surface of the
titanium stem (Willert et al., 1996). Despite mixed reviews about the use of cemented
titanium in femoral implants, it is still commonly employed in cemented upper-limb
implant applications (Van der Lugt and Rozing, 2004).
21
1.3 IMPLANT STABILITY
Implant stability is a term used to describe the ability of a joint replacement system
to resist movement and loosening. In the case of cemented implants, stability can be
measured (in part) by quantifying the displacement, or micromotion of the implant stem
relative to the surrounding cement.
A stable implant system incorporates a stem design that utilizes optimal geometry,
surface mechanics, and material properties that facilitate transmission of loads to the
surrounding cement and bone, without creating damaging stresses and excessive stem
motion (Scheerlinck and Casteleyn, 2006). To achieve stability of implant systems, two
methods of implant design are explored to ensure mechanical longevity over repeated
loading; force-closed and shape-closed designs (Huiskes et al., 1998).
Force closed systems, such as straight, polished stem designs, are aimed to allow
initial stem subsidence during interface loading (Huiskes et al., 1998). Subsidence of the
implant stem after interface debonding generates frictional forces at the stem-cement
interface, which balances the applied load. This equilibrium of forces at the interface,
therefore, creates a stable implant condition. In comparison, shape-closed systems such
as longitudinally curved and non-circular cross section stems, as well as modified stem
surface designs (i.e., surface treatments and finishes), take into consideration initial
stability of the implant system. This is done by ensuring a well fixed stem-cement
system during fixation, so that migration of the stem does not occur. This method aims to
guarantee no motion at the stem interface, therefore preventing the initiation of loosening
(Huiskes et al., 1998). Although some force closed systems have been successful for
particular joint replacement designs, this thesis will focus on shape closed systems with a
particular interest in surface treatments.
22
1.3.1 IMPLANT STABILITY MEASUREMENTS IN-VITRO
Clinically, implant migration is observed from radiographic analysis of the stem-
cement interface, particularly regions of radiolucency surrounding the implant stem
(Chambers et al., 2001), with regions greater than 1 mm indicative of definite loosening
(Banaszkiewicz, 2009b). Experimentally, there are a variety of tools used to measure
implant micromotion.
Linear variable differential transducers (LVDT’s) are used to measure displacement
of the implant stems in-vitro. By drilling window holes through the bone and/or cement
mantle, the LVDT or pins connected to the LVDT, can be inserted through the holes to be
in direct contact with the implant stem (DiSilvestro et al., 2004; Doehring et al., 1999;
Engh et al., 1992; Maher et al., 2001). Once there is displacement of the stem, this is
recorded by the LVDT as a voltage change and resulting micromotion measurement.
This technique has proven useful, however, the method is invasive and LVDT’s are
specific to measuring linear displacement only, so therefore not very effective when
rotational motion is of interest.
Radiostereophotogrammetric analysis is another method used for implant
displacement measurement both in-vivo and in-vitro. Metal bead markers are placed onto
the bone, cement and implant during the replacement procedure, and motion at the
interfaces can be measured from stereoradiograph images taken before and after loading
(Nilsson et al., 1991). Although this method offers information about motion occurring at
the interface, it can only measure displacement of the implant once loading is complete.
This can be problematic for studies interested in investigating how loading affects
micromotion in real time.
Another technique incorporates the use of an optical system with reflective markers
to observe motion of the implant relative to the bone (Westphal et al., 2006). Markers
are placed on the implant and bone, and cameras are used to track the motion of the
markers throughout mechanical tests. The software associated with optical systems can
then determine the relative distances of the implant and bone markers, to obtain
measurements of implant stability. This method, however, may not be suitable for
23
measuring motion at the level of the stem cement-cement interface, since optical markers
associated with these systems may be limited by the size and volume restrictions at the
cemented interface.
More recently, stability studies done by Mann et al., and Race et al., discussed
digital image correlation (DIC) as a method for measuring micromotion at the stem-
cement and cement-bone interfaces (Mann et al., 2010; Race et al., 2010). This method
involved using a camera with telecentric lens, and custom written software to document
motion occurring at the stem, cement and bone interfaces during torsional loading of
transverse sections of the implanted stem. The optical system (i.e., camera and
telecentric lens) recorded video of the interfaces during mechanical loading, and the
software program measured relative displacement of sampling locations placed on either
side of the interfaces. While this methodology was only useful for detecting interface
motion of sectioned implant stems during torsional loading, the optical system served as
an imaging modality for observing the interfaces. As such, a similar system could be
adapted for use in future interface stability studies with careful selection of the required
optical components (see below).
1.4 OPTICAL SYSTEMS
As mentioned previously, optical systems can be useful as a tool for measuring
interface motion. Depending on the environment of implant testing, as well as the
resolution of the intended measurement, the hardware and software components of the
optical system can be carefully chosen to ensure best optical system design. While the
software components are specific to the optical system, the hardware components are
generally standard, and include a camera system (i.e., camera with lens) and source of
illumination (Relf, 2004).
24
1.4.1 CAMERA SYSTEMS
Like all digital based cameras, those used in optical systems operate on a principle,
where reflected light from the object being imaged enters the camera through the lens,
and is focused and transmitted to the camera sensor (Figure 1.8) (Relf, 2004). The
camera’s sensor is made up of an array of sensors called pixel elements (i.e., pixels),
which act as sites for the collection of the transmitted light. One type of sensor used in
cameras is known as a charged couple device (CCD), where the collected light photons in
each of the sensors are converted to electric charges (Relf, 2004). The produced charge
is directly proportional to the number of light photons that hit the sensor. The charge is
then converted to voltage that is read by a digital converter as a range of integers, which
represents the location and brightness value for each element of the sensor array. These
brightness values can range between 0–255 in the red, green and blue channels that make
up the image’s color (Nakamura, 2005).
The camera system’s resolution is the measure of detail that can be discerned in an
image (see Relf, 2004) (Figure 1.8). For measurement applications, the resolution of the
camera system should be appropriate in order to detect changes in the image. Camera
system resolution is influenced by its sensor pixel size, in addition to the characteristics
of the lens (Nakamura, 2005). For a good system resolution, the lens must be compatible
with the sensor, facilitating optimal diffraction of light relative to the sensor’s pixel size.
This characteristic of the lens is controlled by its f-number, which is a ratio of the lens
focal length to the diameter of its aperture (Relf, 2004).
In addition to the system resolution, the characteristics of the lens also influence the
magnification changes with working distance (Figure 1.8). Typical lenses show greater
magnification for shorter working distances (i.e., the object is closer to the lens). This is
not useful for optical measurement systems, since it gives a false representation of object
sizes in the image’s field of view. As such, telecentric lens have been introduced to
mitigate the dependency of object distance on image magnification, by controlling the
path of the rays entering the optical system.
25
With a telecentric lens, the magnification of the object is independent of working
distance, providing the object stays within the lens’ depth of field (see Relf, 2004)
(Figure 1.8). Telecentric lens are designed with specific features in order to limit the type
of rays entering the optical system. Only light rays that are parallel to the main axis, or
have a principle angle of zero, are collected by the lens. For a conventional lens, this
principle ray angle can be quite large depending on the distance of the object from the
lens, resulting in large changes in image magnification (Figure 1.9). However, by
limiting the size of the principle angle, the magnification errors can be reduced. In
addition to magnification control, high quality telecentric lens show low degrees of image
distortion and reduced perspective errors.
1.4.2 ILLUMINATION
The source of illumination is another important hardware component of the optical
system. The basis of optical systems depends on the harvesting of light to form images.
In particular, the harvesting of light for optical systems used in experimental design needs
to be of a controlled nature, to reduce effects of variability in resulting image
measurements (Relf, 2004). By determining an appropriate source of illumination for a
particular optical system, light rays can be controlled to fall within the image’s field of
view, reducing the effects of glare.
When choosing an appropriate light source, the type of reflection expected from the
object needs to be considered. There are two main types of reflections expected; specular
and diffuse (Relf, 2004). Specular reflection is bright, and occurs in a single direction,
such as the reflection from smooth, shiny surfaces. This type of reflection is variable
because it disappears with change in positioning of the illuminator or object. Specular
surfaces are best lit with diffused illumination, which allows optical imaging without
bright reflections. In comparison, diffuse reflection is quite faint but not variable, and
occurs in many directions. Reflection from rough or textured surfaces is an example of
diffuse reflection.
26
Figure 1.8: Schematic of a Typical Camera System
The optical system consists of a lens positioned at a working distance from the object
being imaged, through which reflected light is focused onto the camera sensor, and
produces a resultant image. The distance between the lens and the camera sensor is a
measure of the focal length (f). The overall size of object space is known as the Field of
View (FOV). For any two objects located within this field of view, the resolution of the
camera system dictates how far apart the objects can be, and still be perceived as
individual entities. The depth of field represents the range of the working distance in
which the objects are in focus.
27
Figure 1.9: Schematic of a Conventional Lens versus Telecentric Lens
For a conventional lens, the magnification of the image is dictated by the distance of the
object from the lens. An object situated close to the lens would appear larger due to the
larger angle of the principle ray entering the lens. In comparison, a telecentric lens limits
the type of rays entering the lens, filtering only the light rays that are parallel to the main
axis. As a result, the principle rays are maintained at angles close to 0º, producing an
image with magnification independent of the objects’ distance from the lens.
28
1.5 COMPUTED TOMOGRAPHY (CT) IMAGING
Computed tomography (CT) has been used in the field of orthopaedics as a tool
for studying bones and joints in a three-dimensional space. In particular, CT imaging has
been used to investigate fracture biomechanics of joints, as well as joint contact
mechanics (Greenspan, 2011; Lalone, 2012). With regards to the biomechanics of joint
replacement systems, CT analysis has been less popular due to the sensitivity of the
system to metal artifacts. Metal artifacts can produce bands or streaks across the image,
causing loss of visualization, as well as image data. The effect that metal has on CT
images can be explained by the operations of the CT system.
CT systems work on the basis of x-ray imaging, where x-rays emitted from a
source interact with an object, and results in absorbance, scatter, and transmission of rays
(Stock, 2009). With CT imaging, the source produces a narrow, fan-shaped beam of x-
rays which rotate around the object being imaged, and detectors on the exit side of the
object receive the transmitted rays. During one rotation of the source, multiple x-rays are
transmitted through the object at various angles. Depending on the density value of the
object being irradiated, the amount of x-rays absorbed or scattered varies, causing
reduced intensity (i.e., attenuation) of the transmitted rays. The attenuation is measured
by the ratio of absorbed or scattered rays per unit thickness of the object, and this value is
used to determine the corresponding Hounsfield Unit (HU) or CT number of the object.
The attenuated rays collected by the detector are represented as individual images at the
various angles of x-ray projection, which are stitched together to create a cross-sectional
image of the object. When the object is translated through the rotating x-ray source,
multiple cross-sectional images are collected along the translational length, and are used
to create a three-dimensional model of the object being imaged.
With regards to metals, however, the density values and resulting CT numbers are
beyond the normal range that can be handled by the CT software, causing incomplete
attenuation profiles and resulting metal artifacts (Barrett and Keat, 2004). Metal artifacts
are more prominent in higher atomic number metals such as stainless steel and cobalt
alloys, but are less prominent with lower atomic number metals such as titanium (Boas
29
and Fleischmann, 2012). CT imaging is more likely to be affected by metal artifacts
compared to conventional radiographs due to the larger number of detector measurements
obtained in a single scan (Barrett and Keat, 2004). However, techniques have been
proposed to reduce metal artifacts in CT images (Boas and Fleischmann, 2011; Olsen et
al., 2000; Wang et al., 2000).
In addition to artifact limitations of joint replacement systems in CT imaging,
resolution of the CT system has not been useful for the study of joint replacement
biomechanics. Conventional CT scanners provide a resolution on the order of 1–2 mm
(Kalender, 1995), however, instability of implant systems is detected by micromotion
occurring at the stem-cement interface. This can potentially be overcome through the use
of micro-computed tomography (µCT). As a high resolution CT technology, voxel sizes
of µ-CT systems can vary between 50–250 μm, depending on the manufacturer and
specifications (Stock, 2009). µ-CT systems differ from conventional systems with
regards to their smaller field-of-view, as well as higher resolution detector. These
changes allow for more detailed models of micro-architecture. As such, µ-CT analysis
has been used orthopaedics to quantify implant wear (Bowden et al., 2005; Teeter, 2012),
as well as for the analysis of bone micro-architecture (Waarsing et al., 2005). However,
its use in joint replacement biomechanics has not been well-established. With new
methods being developed to improve on metal artifact limitations, µ-CT imaging may
prove useful as a tool for assessing implant loosening micromechanics at the stem-cement
interface.
1.6 STUDY RATIONALE
The prevalence of joint replacement surgeries has increased over the last decade,
and numbers are expected to continue to increase as a result of the treatment of
musculoskeletal conditions (degenerative joint diseases and joint injury) with the aging of
the baby boomer population (Kurtz et al., 2009; Perruccio et al., 2006). Within recent
30
years, joint replacement surgeries alone accounted for 25% of all orthopedic surgeries in
Ontario (Canizares et al., 2009).
This demand for joint replacement surgery, however, is not limited to the older
population, with reports showing an increase in total joint replacements performed on
younger patients or patients less than 65 years old (Kurtz et al., 2009). In addition,
because of increasing rates of obesity, arthritis and expanded clinical criteria for
eligibility, the need for joint replacement surgery has become overwhelming (Gelber et
al., 1999; Karlson et al., 2003).
The rise in popularity for primary joint replacement surgeries is also expected to
trigger the increase in revision joint replacement surgeries. Although joint replacements
are successful, failure of the replacement systems can result in revision surgeries (Clohisy
et al., 2004). Taking this into consideration, it seems necessary to explore ways of
improving the longevity of these implant systems.
This may be realized, in part, through the alteration in the design of implant stems.
Stem surface modification is an example of stem design that can contribute to reducing
the onset of implant loosening. Stem surface finish has been explored for improving
cemented implant stability (Datir et al., 2006; Huiskes et al., 1998; Jamali et al., 2006),
however, the use of stem surface treatments with cemented implants is one area of stem
design that has not been widely investigated. A clinical study by Jeon et al., addressed
the success of surface treated stems (i.e., plasma sprayed and beaded) on the survival
rates of cemented elbow systems (Jeon et al., 2012), but there are no known experimental
studies that have assessed these surface treatments for improved implant stability. While
these treatments are currently only employed with cemented upper limb replacement
designs, knowledge regarding their stability response could also be applied to the design
of other joint replacement systems.
Titanium and cobalt chrome are two common metals used in implant stem design
(Korkusuz and Korkusuz, 2004; Navarro et al., 2008). However, there has been varied
clinical success with the use of titanium and cobalt chrome stem substrate materials with
surface treated implants, where increased rates of fracture have been reported for sintered
31
bead treatments on titanium components (Athwal and Morrey, 2006; Jeon et al., 2012).
In addition to the variability with surface treated stems, there has also been mixed
reviews over the use of cemented titanium stems in lower limb joint. In particular,
studies have reported increased rates of loosening, and high sensitivity to corrosion,
associated with the femoral component at the hip joint (Hallam et al., 2004; Schweizer et
al., 2005). Despite these varied clinical outcomes with titanium stems, they are still
commonly used in upper limb joints (Van der Lugt and Rozing, 2004), and as such, are
important to consider in stability studies of cemented implant systems.
The stability of implant systems is directly affected by joint loading. Joint loads
acting on the implant are transmitted to the stem-cement interface. If the fixation
strength at the interface is less than that of the acting loads, the stem can experience
debonding and eventual loosening. Therefore, it is necessary to explore the effects of
joint loading in implant stability studies.
Taking these factors into consideration, the overall goal of this thesis was to
determine the role of stem surface treatment and substrate material on the stem-cement
interface stability of implanted stems under various loading modes. Studies presented
within this thesis focused on surface treatments used clinically with upper limb
replacement systems; however, their role in cemented implant stability was analyzed for
generic joint replacement applications, as well as elbow specific components. With the
knowledge gained from this research, it is expected that improvements can be made to
the surface design of current implant stems. Improved implant stem design will enhance
the longevity of implant systems by enhancing the stem’s function as an implant
stabilizer, thus reducing the implant’s rate of loosening. Ultimately, this will improve
patient care by decreasing the chances of revision surgery after a joint replacement
procedure, and consequently reduce the number of joint replacement surgeries, and
subsequent costs to the healthcare system.
32
1.7 SPECIFIC OBJECTIVES AND HYPOTHESES
The specific objectives of this work were as follows, to:
1. develop and validate an optical tracking system to quantify stem-cement interface
micromotion in stemmed implant components;
2. experimentally investigate the effect of stem surface treatment on the torsional
stability of titanium and cobalt chrome stems;
3. compare the roles of stem surface treatment and substrate material on the stability
of implant stem subjected to cyclic compressive loading;
4. determine the contribution of stem circumferential grooves to the stability of
cemented stemmed joint replacement systems under compression and torsional
loading;
5. compare the stability response of stem surface treatment and substrate material
during the application of a bending moment; and
6. use µ-CT imaging to visualize the stem-cement interface during bending, and
thereby investigate the influence of bending loads on internal stem-cement
interface motion.
The corresponding hypotheses were as follows:
1. A custom optical tracking system could be developed to quantify stem-cement
interface motion, on the order of 5 µm in stemmed implant systems, and this
system could be used to non-invasively analyze interface motion.
2. Surface treated implant stems would provide improved torsional stability
compared to smooth surfaces, with cobalt chrome stems showing less rotation in
torsion than titanium stems, due to its higher shear modulus.
33
3. Smooth stems would provide the least resistance to cyclic compressive loading
compared to the surface treated stems, and there would be no effect of stem
material on implant stability under compression.
4. Application of circumferential grooves would not affect torsional stability of
cemented stemmed joint replacement systems, but would have a stabilizing
influence in compression.
5. Smooth stem surfaces would show greatest instability under bending, with
titanium stems experiencing greater interface motion than cobalt chrome stems
due to a lower flexural modulus of elasticity.
6. Micro-CT analysis of cemented stems would show that motion along the internal
length of the stem-cement interface during bending is influenced by creep of the
surrounding bone cement, causing the implant to experience tilting without
apparent loosening.
1.8 THESIS OVERVIEW
This thesis is written in an integrated article format, with each of the above
objectives corresponding to a chapter of the thesis.
Chapter 2 describes the development and validation of the optical tracking system
used for measuring stem-cement interface micromotion in subsequent chapters of this
thesis. The chapter outlines the resolution, accuracy and reliability of the tracking system
to reproduce motion applied by a micrometer gauge.
Chapter 3 investigates the role of clinically relevant surface treatments (i.e.,
smooth, beaded and plasma spray) on the torsional stability of titanium and cobalt
chrome stems. The failure torques and interface stability were analyzed for each of the
surface treated stems under simulated joint torsional loads.
34
Chapter 4 investigates the effect of stem surface treatments and material on the
pistoning failure scenario of implant stems. Smooth, beaded and plasma spray surface
treatments applied to cobalt chrome and titanium stems were compared for differences in
interface strength and interface stability, under cyclic compression loads.
Chapter 5 investigates the effect of a new stem surface design, circumferential
grooves, on the stability of cemented stems under compression and torsional loading.
Applying the observations and results from Chapters 3 and 4, an alternate surface
morphology was investigated to address the failure mechanisms associated with surface
treated stems.
Chapter 6 investigates the effect of stem surface treatments and material on
implant stability under bending loads. The interface motion for each of the surface
treated stems was observed using the optical tracking system described in Chapter 2, and
the contribution of the cement mantle in interface stability was addressed.
Chapter 7 uses µ-CT imaging to visualize and quantify motion at the stem-cement
interface with an applied bending moment. Based on the results from Chapter 6, it was
determined that a technique to visualize internal stem- cement motion was needed to
understand the contribution of bending moments to interface stability. Micro-CT
measurements, made along the length of the stem-cement interface, were used to
investigate the contribution of the stem and cement to interface motion during the
application of real-time bending loads.
Chapter 8 is the concluding chapter of this thesis, and outlines the main findings of
the individual studies detailed within this thesis, along with their strengths and
limitations, as well as offers some suggestions for future work in the area of implant
interface biomechanics.
35
1.9 REFERENCES
Amis, A.A., 2012. Biomechanics of the Elbow. In: Stanley, D., Trail, I. (Eds.), Operative
Elbow Surgery. Churchill Livingstone Elsevier, Edinburgh; New York, pp. 29–44.
Amis, A.A., Dowson, D., Wright, V., 1980. Elbow joint force predictions for some
strenuous isometric actions. Journal of Biomechanics 13, 765–775.
Athwal, G.S., Morrey, B.F., 2006. Revision total elbow arthroplasty for prosthetic
fractures. The Journal of Bone and Joint Surgery; American Volume 88, 2017–2026.
Austman, R.L., Beaton, B.J., Quenneville, C.E., King, G.J., Gordon, K.D., Dunning,
C.E., 2007. The effect of distal ulnar implant stem material and length on bone
strains. Journal of Hand Surgery 32, 848–854.
Australian Orthopaedic Association, T., 2010. National Joint Registry, Hip and Knee
Arthroplasty Annual Report 2010,
http://www.dmac.adelaide.edu.au/aoanjrr/documents/aoanjrrreport_2010.pdf.
Banaszkiewicz, P., 2009a. The basic science oral. In: Banaszkiewicz, P., Kader, D.,
Maffulli, N. (Eds.), Graduate Orthopaedics. Cambridge University Press, New York,
pp. 514–516.
Banaszkiewicz, P., 2009b. Hip oral core topics. In: Banaazliewicz, P., Kader, D.,
Maffulli, N. (Eds.), Postgraduate Orthopaedics. Cambridge University Press, New
York, pp. 155–193.
Barrack, R.L., 2000. Early failure of modern cemented stems. The Journal of
Arthroplasty 15, 1036–1050.
Barrett, J., Keat, N., 2004. Artifacts in CT: Recognition and avoidance. Radiographics
1679–1691.
Bauer, T.W., Schils, J., 1999. The pathology of total joint arthroplasty. I. Mechanisms of
implant fixation. Skeletal Radiology 28, 423–432.
36
Beksac, B., Taveras, N.A., Valle, A.G., Salvati, E.A., 2006. Surface finish mechanics
explain different clinical survivorship of cemented femoral stems for total hip
arthroplasty. Journal of Long-term Effects of Medical Implants 16, 407–422.
Bergmann, G., 1987. Biomechanics and pathomechanics of the shoulder joint with
reference to prosthetic joint replacement. In: Kolbel, R., Helbig, B., Blauth, W.
(Eds.), Shoulder Replacement. Springer-Verlag, Berlin, pp. 33–43.
Bergmann, G., Deuretzbacher, G., Heller, M., Graichen, F., Rohlmann, A., Strauss, J.,
Duda, G.N., 2001. Hip contact forces and gait patterns from routine activities.
Journal of Biomechanics 34, 859–871.
Bergmann, G., Graichen, F., Bender, A., Kääb, M., Rohlmann, A., Westerhoff, P., 2007.
In vivo glenohumeral contact forces--measurements in the first patient 7 months
postoperatively. Journal of Biomechanics 40, 2139–2149.
Bergmann, G., Graichen, F., Rohlmann, A., 1993. Hip joint loading during walking and
running, measured in two patients. Journal of Biomechanics 26, 969–990.
Bergmann, G., Graichen, F., Rohlmann, A., 2004. Hip joint contact forces during
stumbling. Langenbecks Archives of Surgery 389, 53–59.
Berzins, A., Sumner, D.R., Andriacchi, T.P., Galante, J.O., 1993. Stem curvature and
load angle influence the initial relative bone-implant motion of cementless femoral
stems. Journal of Orthopaedic Research 11, 758–769.
Boas, F., Fleischmann, D., 2011. Evaluation of two iterative techniques for reducing
metal artifacts in computed tomography. Radiology 259, 894–902.
Boas, F.E., Fleischmann, D., 2012. CT artifacts: causes and reduction techniques.
Imaging in Medicine 4, 229–240.
Bobyn, J.D., Pilliar, R.M., Cameron, H.U., Weatherly, G.C., 1980. The optimum pore
size for the fixation of porous-surfaced metal implants by the ingrowth of bone.
Clinical Orthopaedics and Related Research 150, 263–270.
37
Bowden, a E., Kurtz, S.M., Edidin, a a, 2005. Validation of a micro-CT technique for
measuring volumetric wear in retrieved acetabular liners. Journal of Biomedical
Materials Research. Part B, Applied Biomaterials 75, 205–209.
Boyer, P., Lazennec, J.Y., Poupon, J., Rousseau, M.A., Ravaud, P., Catonne, Y., 2009.
Clinical and biological assessment of cemented titanium femoral stems: an 11-year
experience. International Orthopaedics 33, 1209–1215.
Brånemark, R., Brånemark, P.I., Rydevik, B., Myers, R.R., 2001. Osseointegration in
skeletal reconstruction and rehabilitation: a review. Journal of Rehabilitation
Research and Development 38, 175–181.
Buechel, F.F., Pappas, M.J., SpringerLink, 2012. Properties of Materials in Orthopaedic
Implant Systems. In: Principles of Human Joint Replacement. Springer Berlin
Heidelberg, Berlin, Heidelberg, pp. 1–35.
Bundy, K.J., Penn, R.W., 1987. The effect of surface preparation on metal/bone cement
interfacial strength. Journal of Biomedical Materials Research 21, 773–805.
Callaghan, J.J., Fulghum, C.S., Glisson, R.R., Stranne, S.K., 1992. The effect of femoral
stem geometry on interface motion in uncemented porous-coated total hip
prostheses. Comparison of straight-stem and curved-stem designs. The Journal of
Bone and Joint Surgery; American Volume 74, 839–848.
Canizares, M., MacKay, C., Davis, A.M., Mahomed, N., Badley, E.M., 2009. A
population-based study of ambulatory and surgical services provided by orthopaedic
surgeons for musculoskeletal conditions. BMC Health Services Research 9, 1–9.
Chambers, I.R., Fender, D., McCaskie, a W., Reeves, B.C., Gregg, P.J., 2001.
Radiological features predictive of aseptic loosening in cemented Charnley femoral
stems. The Journal of Bone and Joint Surgery; British Volume 83, 838–842.
Cheung, E. V, O’Driscoll, S.W., 2007. Total elbow prosthesis loosening caused by ulnar
component pistoning. The Journal of Bone and Joint Surgery; American Volume 89,
1269–1274.
38
Clohisy, J.C., Calvert, G., Tull, F., McDonald, D., Maloney, W.J., 2004. Reasons for
revision hip surgery: a retrospective review. Clinical Orthopaedics and Related
Research 429, 188–192.
Crowninshield, R.D., Brand, R.A., Johnston, R.C., Milroy, J.C., 2006. The effect of
femoral stem cross-sectional geometry on cement stresses in total hip reconstruction.
Clinical Orthopaedics and Related Research 146, 71–77.
Crowninshield, R.D., Jennings, J.D., Laurent, M.L., Maloney, W.J., 1998. Cemented
femoral component surface finish mechanics. Clinical Orthopaedics and Related
Research 355, 90–102.
Dall, G.F., Simpson, P.M., Mackenzie, S.P., Breusch, S.J., 2007. Inter- and intra-batch
variability in the handling characteristics and viscosity of commonly used antibiotic-
loaded bone cements. Acta Orthopaedica 78, 412–420.
Daly, C., Carey, G., Beverland, D.E., Orr, J.F., Dunne, N.J., 2004. Mixing of acrylic bone
cement - Current theatre practices. In: 7th World Biomaterials Conference.
Biomaterials 2004 Congress Managers, Sydney, p. 1465.
Datir, S.P., Kurta, I.C., Wynn-Jones, C.H., 2006. Ten-year survivorship of rough-
surfaced femoral stem with geometry similar to Charnley femoral stem. The Journal
of Arthroplasty 21, 392–397.
Davies, J.P., Harris, W.H., 1993. Strength of cement-metal interfaces in fatigue:
comparison of smooth, porous and precoated specimens. Clinical Materials 12, 121–
126.
Davis, J.R., 2003. Coatings. In: Handbook of Materials for Medical Devices. ASM
International, Materials Park, OH, pp. 179–194.
DiSilvestro, M.R., Sherman, J.T., Dietz, T.L., 2004. A new position measurement system
for micromeasurements in orthopaedics. In: Engineering in Medicine and Biology
Society, 26th Annual Conference of the IEEE. IEEE, San Francisco, pp. 2438–2441.
39
Doehring, T.C., Rubash, H.E., Dore, D.E., 1999. Micromotion measurements with hip
center and modular neck length alterations. Clinical Orthopaedics and Related
Research 362, 230–239.
Einsiedel, T., Gebhard, F., Bregolato, I., Hiemeier, A., Kinzl, L., Schultheiss, M., 2008.
Proximal cement fixation in total hip arthroplasty--first results with a new stem
design. International Orthopaedics 32, 295–306.
Engh, C.A., Bobyn, J.D., 1988. The influence of stem size and extent of porous coating
on femoral bone resorption after primary cementless hip arthroplasty. Clinical
Orthopaedics and Related Research 231, 7–28.
Engh, C.A., O’Connor, D., Jasty, M., McGovern, T.F., Bobyn, J.D., Harris, W.H., 1992.
Quantification of implant micromotion, strain shielding, and bone resorption with
porous-coated anatomic medullary locking femoral prostheses. Clinical
Orthopaedics and Related Research 285, 13–29.
Evans, B.G., Daniels, A.U., Serbousek, J.C., Mann, R.J., 1988. A comparison of the
mechanical designs of articulating total elbow prostheses. Clinical Materials 3, 235–
248.
Geesink, R.G., De Groot, K., Klein, C.P., 1988. Bonding of bone to apatite-coated
implants. The Journal of Bone and Joint Surgery; British Volume 70, 17–22.
Geiger, M.H., Keating, E.M., Ritter, M.A., Ginther, J.A., Faris, P.M., Meding, J.B., 2001.
The clinical significance of vacuum mixing bone cement. Clinical Orthopaedics and
Related Resarch 382, 258–266.
Gelber, A.C., Hochberg, M.C., Mead, L.A., Wang, N.Y., Wigley, F.M., Klag, M.J., 1999.
Body mass index in young men and the risk of subsequent knee and hip
osteoarthritis. The American Journal of Medicine 107, 542–548.
Glass, L., Pierfrancesco, R., 2011. Macromanagement: Surface Manufacturing for
Implants Orthotec 2, 4–8.
40
Greenspan, A., 2011. Imaging Techniques in Orthopedics. In: McAllister, L., Barrett, K.,
Rampertab, R. (Eds.), Orthopedic Imaging: A Practical Approach. Lippincott
Williams & Wilkins, Philadelphia, PA, pp. 17–38.
Guerra, S.M., 2004. Design and development of a total elbow prosthesis to quantify
ulnohumeral load transfer. (M.E.Sc Thesis) University of Western Ontario, pp 85-
108.
Hallab, N.J., Urban, R.M., Jacobs, J.J., 2004. Corrosion and Biocompatibility of
Orthopedic Implants. In: Yaszemski, M., Trantolo, D., Lewandrowski, K., Hasirci,
V., Altobelli, D., Wise, D. (Eds.), Biomaterials in Orthopedics. Marcel Dekker, Inc.,
New York, pp. 63–91.
Hallam, P., Haddad, F., Cobb, J., 2004. Pain in the well-fixed, aseptic titanium hip
replacement. The role of corrosion. The Journal of Bone and Joint Surgery; British
Volume 86, 27–30.
Hansen, D., Jensen, J.S., 1990. Prechilling and vacuum mixing not suitable for all bone
cements. Handling characteristics and exotherms of bone cements. The Journal of
Arthroplasty 5, 287–290.
Harris, L.J., Tarr, R.R., 1979. Implant failures in orthopaedic surgery. Biomaterials
Medical Devices And Artificial Organs 7, 243–255.
Hernigou, P., Daltro, G., Lachaniette, C.H., Roussignol, X., Mukasa, M.M., Poignard, A.,
2009. Fixation of the cemented stem: clinical relevance of the porosity and thickness
of the cement mantle. Open Orthopaedic Journal 3, 8–13.
Hinrichs, F., Kuhl, M., Boudriot, U., Griss, P., 2003. A comparative clinical outcome
evaluation of smooth (10-13 year results) versus rough surface finish (5-8 year
results) in an otherwise identically designed cemented titanium alloy stem. Archives
of Orthopaedic and Trauma Surgery 123, 268–72.
Hosein, Y.K., Clynick, M.P., McLachlin, S.D., King, G.J.W., Dunning, C.E., 2012. The
effect of stem curvature on torsional stability of a generalized cemented joint
41
replacement system. Journal of Applied Biomaterials & Functional Materials doi:
10.5301/JABFM.2012.9266.
Huiskes, R., Verdonschot, N., Nivbrant, B., 1998. Migration, stem shape, and surface
finish in cemented total hip arthroplasty. Clinical Orthopaedics and Related
Research 355, 103–112.
Iesaka, K., Jaffe, W., Kummer, F., 2004. Effects of the initial temperature of acrylic bone
cement liquid monomer on the properties of the stem–cement interface and cement
polymerization. Journal of Biomedical Materials Research Part B, Applied
Biomaterials 68, 186–190.
Incavo, S.J., Havener, T., Benson, E., McGrory, B.J., Coughlin, K.M., Beynnon, B.D.,
2004. Efforts to improve cementless femoral stems in THR: 2- to 5-year follow-up
of a high-offset femoral stem with distal stem modification (Secur-Fit Plus). The
Journal of Arthroplasty 19, 61–67.
Jamali, A.A., Lozynsky, A.J., Harris, W.H., 2006. The effect of surface finish and of
vertical ribs on the stability of a cemented femoral stem: an in vitro stair climbing
test. The Journal of Arthroplasty 21, 122–128.
Janssen, D., Mann, K.A., Verdonschot, N., 2008. Micro-mechanical modeling of the
cement-bone interface: the effect of friction, morphology and material properties on
the micromechanical response. Journal of Biomechanics 41, 3158–3163.
Jeon, I.H., Morrey, B.F., Sanchez-Sotelo, J., 2012. Ulnar component surface finish
influenced the outcome of primary Coonrad-Morrey total elbow arthroplasty.
Journal of Shoulder and Elbow Surgery 21, 1229–1235.
Johnson, J.A., King, G.J.W., 2005. Anatomy and Biomechanics of the Elbow. In:
Williams, G.R. (Ed.), Shoulder and Elbow Arthroplasty. Lippincott Williams &
Wilkins, Philadelphia, pp. 279–296.
Kalender, A., 1995. Thin-Section Three dimensional Spiral CT: Is Isotpropic Imaging
Possible. Radiology 197, 578–580.
42
Karlson, E.W., Mandl, L.A., Aweh, G.N., Sangha, O., Liang, M.H., Grodstein, F., 2003.
Total hip replacement due to osteoarthritis: the importance of age, obesity, and other
modifiable risk factors. The American Journal of Medicine 114, 93–98.
Kedgley, A.E., Takaki, S.E., Lang, P., Dunning, C.E., 2007. The effect of cross-sectional
stem shape on the torsional stability of cemented implant components. Journal of
Biomechanical Engineering 129, 310–314.
Korkusuz, P., Korkusuz, F., 2004. Hard Tissue-Biomaterial Interactions. In: Yaszemski,
M.J., Trantolo, D.J., Lewandrowski, K.-U., Hasirci, V., Altobelli, D., Wise, D.
(Eds.), Biomaterials in Orthopaedics. Marcel Dekker, Inc., New York, pp. 2–6.
Kuhn, K.-D., 2009. Properties of Bone Cement: What is Bone Cement? In: Breusch, S.J.,
Malchau, H. (Eds.), The Well-cemented Total Hip Arthroplasty: Theory and
Practice. Springer Berlin Heidelberg, New York, pp. 52–59.
Kurtz, S.M., Lau, E., Ong, K., Zhao, K., Kelly, M., Bozic, K.J., 2009. Future Young
Patient Demand for Primary and Revision Joint Replacement: National Projections
from 2010 to 2030. Clinical Orthopaedics and Related Research 467, 2606–2612.
Kutzner, I., Heinlein, B., Graichen, F., Bender, A., Rohlmann, A., Halder, A., Beier, A.,
Bergmann, G., 2010. Loading of the knee joint during activities of daily living
measured in vivo in five subjects. Journal of Biomechanics 43, 2164–2173.
Lalone, E., 2012. An Image-Based Tool to Examine Joint Congruency at the Elbow.
(Ph.D Thesis) The University of Western Ontario, pp. 1-49.
Lewis, G., 1997. Properties of acrylic bone cement: state of the art review. Journal of
Biomedical Materials Research 38, 155–182.
Macaulay, W., DiGiovanni, C.W., Restrepo, A., Saleh, K.J., Walsh, H., Crossett, L.S.,
Peterson, M.G., Li, S., Salvati, E.A., 2002. Differences in bone-cement porosity by
vacuum mixing, centrifugation, and hand mixing. The Journal of Arthroplasty 17,
569–575.
43
Maher, S.A., Prendergast, P.J., Lyons, C.G., 2001. Measurement of the migration of a
cemented hip prosthesis in an in vitro test. Clinical Biomechanics 16, 307–314.
Mann, K. a, Ayers, D.C., Damron, T. a, 1997. Effects of stem length on mechanics of the
femoral hip component after cemented revision. Journal of Orthopaedic Research
15, 62–8.
Mann, K.A., Miller, M.A., Verdonschot, N., Izant, T.H., Race, A., 2010. Functional
interface micromechanics of 11 en-bloc retrieved cemented femoral hip
replacements. Acta Orthopaedica 81, 308–317.
Maurer, T.B., Ochsner, P.E., Schwarzer, G., Schumacher, M., 2001. Increased loosening
of cemented straight stem prostheses made from titanium alloys. An analysis and
comparison with prostheses made of cobalt-chromium-nickel alloy. International
Orthopaedics 25, 77–80.
Mohler, C.G., Callaghan, J.J., Collis, D.K., Johnston, R.C., 1995. Early loosening of the
femoral component at the cement-prosthesis interface after total hip replacement.
The Journal of Bone and Joint Surgery; American Volume 77, 1315–1322.
Morshed, S., Bozic, K.J., Ries, M.D., Malchau, H., Colford, J.M., 2007. Comparison of
cemented and uncemented fixation in total hip replacement: a meta-analysis. Acta
Orthopaedica 78, 315–26.
Nakamura, J., 2005. Basics of Image Sensors. In: Nakamura, J. (Ed.), Image Sensors and
Signal Processing for Digital Still Cameras. Taylor & Francis Group, Boca Raton,
FL, pp. 53–93.
Navarro, M., Michiardi, A., Castano, O., Planell, J.A., Castaño, O., 2008. Biomaterials in
orthopaedics. Journal of the Royal Society, Interface 5, 1137–1158.
New Zealand Orthopaedic Association, T.N.Z.J.R., 2010. Eleven Year Report- January
1999 to December 2009., http://www.cdhb.govt.nz/njr/reports/A2D65CA3.pdf.
44
Ni, G., Lu, W., Chiu, K., Fong, D., 2005. Cemented or uncemented femoral component
in primary total hip replacement? A review from a clinical and radiological
perspective. Journal Orthopaedic Surgery (Hong Kong) 13, 96–105.
Nilsson, K.G., Karrholm, J., Ekelund, L., Magnusson, P., 1991. Evaluation of
micromotion in cemented vs uncemented knee arthroplasty in osteoarthrosis and
rheumatoid arthritis. Randomized study using roentgen stereophotogrammetric
analysis. The Journal of Arthroplasty 6, 265–278.
Olofsson, K., Digas, G., Karrholm, J., 2006. Influence of design variations on early
migration of a cemented stem in THA. Clinical Orthopaedics and Related Research
448, 67–72.
Olsen, R., Munk, P., Lee, M., 2000. Metal Artifact Reduction Sequence: Early Clinical
Applications. Radiographics 20, 699–712.
Park, J.B., 1992. Orthopedic prosthesis fixation. Annals of Biomedical Engineering 20,
583–94.
Perruccio, A. V, Power, J.D., Badley, E.M., 2006. Revisiting arthritis prevalence
projections--it’s more than just the aging of the population. The Journal of
Rheumatology 33, 1856–1862.
Race, A., Miller, M.A., Mann, K.A., 2010. Novel methods to study functional loading
micromechanics at the stem-cement and cement-bone interface in cemented femoral
hip replacements. Journal of Biomechanics 43, 788–791.
Relf, C., 2004. Setting Up. In: Image Acquisition and Processing with LabVIEW. CRC
Press, Boca Raton, FL, pp. 15–45.
Rodriguez-Gonzalez, F.Á., 2009. Chapter 9: Failure Modes of Implants. In: Biomaterials
in Orthopaedic Surgery. pp. 177–197.
Ryan, G., Pandit, A., Apatsidis, D.P., 2006. Fabrication methods of porous metals for use
in orthopaedic applications. Biomaterials 27, 2651–2670.
45
Saha, S., Pal, S., 1984. Mechanical properties of bone cement: a review. Journal of
Biomedical Materials Research 18, 435–462.
Scheerlinck, T., Casteleyn, P.P., 2006. The design features of cemented femoral hip
implants. The Journal of Bone and Joint Surgery; British Volume 88, 1409–1418.
Schweizer, A., Luem, M., Riede, U., Lindenlaub, P., Ochsner, P.E., 2005. Five-year
results of two cemented hip stem models each made of two different alloys.
Archives of Orthopaedic and Trauma Surgery 125, 80–86.
Serbetci, K., Hasirci, N., 2004. Recent Developments in Bone Cement. In: Yaszemski,
M.J., Trantolo, D.J., Lewandrowski, K.-U., Hasirci, V., Altobelli, D., Wise, D.
(Eds.), Biomaterials in Orthopaedic. Marcel Dekker, Inc., New York, pp. 241–286.
Skinner, H.B., 2006. Basic Science in Orthopaedic Surgery. In: H., S., Naglieri, C., R.Y.,
B. (Eds.), Current Diagnosis and Treatment in Orthopaedics. Lange Medical
Books/McGraw-Hill Companies, pp. 1–63.
Skytta, E.T., Eskelinen, A., Paavolainen, P., Ikavalko, M., Remes, V., 2009. Total elbow
arthroplasty in rheumatoid arthritis: a population-based study from the Finnish
Arthroplasty Register. Acta Orthopaedica 80, 472–477.
Smeds, S., Goertzen, D., Ivarsson, I., 1997. Influence of temperature and vacuum mixing
on bone cement properties. Clinical Orthopaedics and Related Resarch 334, 326–
334.
Stock, S.R., 2009. MicroCT Systems and Their Components. In: Micro Computed
Tomography- Methodology and Applications. Taylor & Francis Group, Boca Raton,
FL, pp. 39–84.
Teeter, M., 2012. Assessment of Wear in Total Knee Arthroplasty Using Advanced
Radiographic Techniques. (Ph.D Thesis) The University of Western Ontario, pp. 1-
36.
46
Thomas, S.R., Shukla, D., Latham, P.D., 2004. Corrosion of cemented titanium femoral
stems. The Journal of Bone and Joint Surgery; British Volume 86, 974–978.
Tortora, G., Nielsen, M., 2012. Synovial Joints. In: Principles of Human Anatomy. John
Wiley & Sons, Inc., Hoboken, NJ, p. 268.
Van der Lugt, J.C., Rozing, P.M., 2004. Systematic review of primary total elbow
prostheses used for the rheumatoid elbow. Clinical Rheumatology 23, 291–298.
Verdonschot, N., 2005. Philosophies of stem designs in cemented total hip replacement.
Orthopedics 28, s833–840.
Verdonschot, N., Huiskes, R., 1995. Dynamic creep behavior of acrylic bone cement.
Journal of Biomedical Materials Research 29, 575–581.
Waarsing, J.H., Day, J.S., Weinans, H., 2005. Longitudinal micro-CT scans to evaluate
bone architecture. Journal of Musculoskeletal & Neuronal Interactions 5, 310–312.
Wang, G., Frei, T., Vannier, M.W., 2000. Fast iterative algorithm for metal artifact
reduction in X-ray CT. Academic Radiology 7, 607–614.
Webb, J.C., Spencer, R.F., 2007. The role of polymethylmethacrylate bone cement in
modern orthopaedic surgery. The Journal of Bone and Joint Surgery; British Volume
89, 851–857.
Westerhoff, P., Graichen, F., Bender, A., Halder, A., Beier, A., Rohlmann, A.,
Bergmann, G., 2009. In vivo measurement of shoulder joint loads during activities
of daily living. Journal of Biomechanics 42, 1840–1849.
Westphal, F.M., Bishop, N., Honl, M., Hille, E., Puschel, K., Morlock, M.M., 2006.
Migration and cyclic motion of a new short-stemmed hip prosthesis--a
biomechanical in vitro study. Clinical Biomechanics 21, 834–840.
47
Willert, H.G., Broback, L.G., Buchhorn, G.H., Jensen, P.H., Koster, G., Lang, I.,
Ochsner, P., Schenk, R., 1996. Crevice corrosion of cemented titanium alloy stems
in total hip replacements. Clinical Orthopaedics and Related Research 333, 51–75.
Wroblewski, B.M., Siney, P.D., Fleming, P.A., 2008. Hip Replacements. In: Sanjukta
Deb (Ed.), Orthopaedic Bone Cements. Woodhead Publishing, Boca Raton, FL, pp.
41–47.
48
CHAPTER 2: DEVELOPMENT AND VALIDATION OF A CUSTOM
OPTICAL TRACKING SYSTEM TO QUANTIFY STEM-CEMENT
INTERFACE MICROMOTION IN STEMMED IMPLANT
COMPONENTS
Overview: Implant loosening is the most common mode of implant failure, which can
begin as small scale micromotion at the implant-cement interface. Quantifying
differences in the magnitude of interface motion is one approach to compare the relative
stability offered by varying implant designs, and serves as a measure of their ability to
resist the effects of loosening. As such, a suitable tool was needed to quantify
micromotion occurring directly at the stem-cement interface. This chapter describes the
development and validation of a custom optical tracking system that will be used for
reliable detection and measurement of interface motion in subsequent chapters.
2.1 INTRODUCTION
Implant micromotion is one measure of implant stability, and can be used in both
clinical and in-vitro studies as a predictor of implant loosening. Most in-vitro studies
have used linear variable differential transducers (LVDT’s), extensometers or sensoring
devices to detect and quantify overall implant motion (Cristofolini et al., 2003; Jamali et
al., 2006; Maher et al., 2001). However, the ability to quantify motion occurring directly
at the stem-cement interface can offer valuable information on implant stability,
eliminating the effects of motion contributed by surrounding fixturing or anatomy. A few
studies have described new methods of investigating interface specific motion (Choi et
al., 2010; Race et al., 2010; Zhang et al., 2009); however, their application involved
invasive methods of accessing the stem-cement interface, in addition to being specific to
loading mode type.
49
Within the field of biomechanics, optical motion capture systems have been used
for defining joint kinematics by tracking the relative motion of landmark markers placed
across the joint (Anglin and Wyss, 2000; Zhou and Hu, 2008). While this method is non-
invasive, optical motion capture systems used for these purposes do not have the system
resolution required for detecting micromotion (Maletskyet al., 2007). However, it was
hypothesized that using appropriate optical equipment and markers, it would be possible
to develop a measurement system capable of non-invasively detecting and measuring
stem-cement interface motion during in-vitro testing.
Therefore, the purpose of this study was to develop and validate an optical
measurement system that incorporated a high resolution camera, appropriately scaled
optical markers, and a custom written data collection and analysis program, to measure
relative micromotion between two moving landmarks in a two-dimensional image space.
2.2 MATERIALS AND METHODS
A pilot GigE Series piA 2400–12gm/gc camera (Basler AG, Ahrensburg,
Germany) with telecentric lens (Opto Engineering, Matua, Italy) and axial diffuse
illuminator (Advanced Illumination, Rochester, VT, USA) comprised the hardware
selected for the optical system (Figure 2.1). These components were carefully chosen to
meet the design requirements of the system; high image resolution with minimal image
distortion or error (Appendix B).
For calibration of the optical system, a grid pattern with 0.1 mm squared grid
intervals (Pyser-SGI Ltd, Kent, UK) was used with a custom written calibration program
(National Instruments Corporation, Austin, TX, USA) (Appendix C.1) to determine the
pixel to millimeter conversion (Figure 2.2). The program incorporated a pixel counting
method, and based on a selected region of the calibration grid chosen by the user, output
the resulting horizontal and vertical pixel count within that region. Using an integrated
scaling method, the pixel to millimeter conversion for both the horizontal and vertical
orientations were determined based on the known intervals of the calibration grid. This
50
conversion was done over nine regions of the grid pattern, which allowed for calibration
over the entire field of view, and an average pixel to mm value was determined and used
for calibration of the optical system. Before the start of the each testing trial, the system
was recalibrated to ensure the accuracy of the resultant measurements.
A custom written LabVIEW program (National Instruments Corporation, Austin,
TX, USA) (Appendix C.2) integrated with the camera capture program (National
Instruments Corporation, Austin, TX, USA), was used for image analysis. This program
used a color thresholding method to detect markers of a specific red-green-blue (RGB)
value in the image capture region, and output a corresponding threshold image (Figure
2.3). The program further incorporated a centroid tracking method to determine the (x,y)
coordinates of the marker centroids in the threshold image window. The data output of
the program included these (x,y) centroid coordinates throughout the video capture
duration, as well as the relative displacement between the markers.
For validation of the optical measuring system, a digit counter inch-micrometer
screw gauge (Fowler Inc., Newton, MA, USA) was used as the gold standard to apply
known displacements. Droplets of fluorescent paint were used as colored markers, and
attached to the anvil and spindle of the micrometer (Figure 2.4). The gauge was moved
through displacements that started at 0.0002 inch (0.005 mm), and increased in
increments of 0.0002 inch (0.005 mm) to a maximum of 0.001 inch (0.025 mm), after
which displacements were set at 0.002 inch (0.051 mm), 0.0039 inch (0.099 mm), 0.0098
inch (0.249 mm), and 0.0197 inch (0.500 mm). This range of displacements was applied
in both the horizontal and vertical directions, over the nine regions of the image field of
view (Figure 2.2). Each displacement was repeated five times, while the optical system
tracked and measured the relative displacement of the markers during each trial.
51
Figure 2.1: Hardware of the Optical Tracking System
A high resolution camera with telecentric lens, and axial diffuse illuminator composed
the hardware of the optical tracking system.
52
Figure 2.2: Calibration Grid used for Pixel to mm Conversion
The calibration grid shown above, with 0.1 mm squared grid spacing, was used for
calibration of the optical system. The exploded view displayed on the far right
shows the grid as seen in the field of view of the optical system. The grid was
divided into nine regions of interest (shown in inset), and pixel to mm conversion
was determined for each of the regions. An average conversion value over all nine
regions was calculated and used for calibration of the optical system.
53
Figure 2.3: Optical Tracking Software
The display of the custom written LabVIEW program used for tracking colored
markers placed on specific landmarks. Shown in the image are the controls used
for setting the color threshold, along with the corresponding camera and threshold
image. Also shown are the (x,y) pixel coordinates of the centroid position of each
marker detected in the camera’s coordinate system.
54
Figure 2.4: Experimental Set-up for Validation of Optical Tracking System
A micrometer screw gauge, with colored markers attached to the anvil and spindle
(shown in the inset image), was used for validation of the optical system. The
micrometer was used as the gold standard to apply known displacements, and the optical
system tracked relative motion of the colored markers.
55
A Bland and Altman plot (Bland and Altman, 1999, 1986; Myles and Cui, 2007)
was used to evaluate agreement between the micrometer gauge and optical system, as
was the percent error measurement. In addition, the reliability of the optical system was
evaluated using the intra-class correlation coefficients (ICC (2,1)) (Shrout and Fleiss,
1979) and Standard Error of Measurements (SEM) of the measured displacements over
the nine regions of the image field of view.
2.3 RESULTS
The Bland and Altman plot showed good agreement between the optical system
and micrometer measurements, based on the scatter of the differences between the
measurements from each system, against the average of the measurements (Figure 2.5).
The scatter of the points generally fell within 1.96 standard deviations of the mean
difference between measurements, with some outliers for measurements greater than
0.1 mm. These outliers were more apparent with 0.5 mm vertical measurements. In
addition, as the average measurement increased, the spread of the data also increased.
This was observed for both the vertical and horizontal displacements, as represented by
the red and black scatter points in Figure 2.5, respectively.
The overall difference based on all of the measured data pooled together between
the micrometer and optical system, as measured by the percent error, was found to be
+8.8% and -9.3% for the displacement output of 0.005 mm to 0.500 mm (Appendix E).
The ICC’s and SEM’s of the vertical and horizontal measurements from the optical
system, over the nine regions of the image field of view are shown Table 2.1. Excellent
reliability was found among the repeated measures from the optical system, where results
of the ICC’s for the vertical and horizontal displacements were all greater than 0.99, and
SEM’s ranged from 0.002 mm to 0.007 mm.
56
Figure 2.5: Bland & Altman Plot Comparing Agreement of Measurement Systems
The Bland and Atman plot showing the difference between the optical system and
micrometer measurements against their average measurements for both vertical and
horizontal displacements. Each point on the plot represents the difference in
measurements (i.e., optical system – micrometer) for the nine applied displacements of
the micrometer, over the nine individual regions of the field of view (Note: To see Bland
and Altman plots for the individual regions, please refer to Appendix D). The plot shows
good agreement between the measurements for all displacements, as most difference
values fall within ±1.96SD of the mean difference. For measurements greater than
0.1 mm, however, there were some outliers that fell outside the region of agreement,
particularly for vertical measurements of 0.5 mm.
57
Table 2.1: Intra-class correlation coefficients (ICC’s) and Standard Errors of
Measurement (SEM) for the Vertical and Horizontal Displacement Measurements
of the Optical System.
Image
Region ICC SEM (mm) ICC SEM (mm)
1 0.9996 0.003 0.9998 0.002
2 0.9976 0.007 0.9998 0.002
3 0.9997 0.003 0.9994 0.004
4 0.9996 0.003 0.9991 0.005
5 0.9996 0.003 0.9997 0.004
6 0.9996 0.003 0.9994 0.004
7 0.9996 0.003 0.9983 0.006
8 0.9987 0.005 0.9994 0.004
9 0.9995 0.003 0.9996 0.003
Vertical Displacement Horizontal Displacement
58
2.4 DISCUSSION
Implant micromotion is a useful measure for defining implant stability. Many
experimental studies involving cemented implants have assessed the survival of implant
systems based on their ability to resist implant micromotion (Burke et al., 1991;
Cristofolini et al., 2003; Maher et al., 2001). These studies, however, assessed implant
motion as a function of stem motion relative to the bone, and not specific to motion
occurring at the stem-cement interface. For cemented implants in particular, it has been
shown that bone cement is susceptible to ‘creep’, where deformation of the cement
occurs under loading conditions (Jeffers et al., 2005; Lewis, 2011; Verdonschot and
Huiskes, 1995). As such, measurement of implant motion at the stem-cement interface
may be a more reliable measure of implant loosening, eliminating the effect of implant
motion from the surrounding cement, as well as the surrounding testing fixtures.
Few studies have reported on stem-cement interface motion as a measure of
implant stability (Choi et al., 2010; Mann et al., 2010; Zhang et al., 2009). Mann et al.,
used a method incorporating digital image correlation to measure displacement along the
interface of sectioned regions of cemented implants under torsional loading (Mann et al.,
2010). In comparison, Zhang et al., described a method which used a custom made
sensor with strain gauges to detect motion occurring between the stem and cement under
compression (Zhang et al., 2009). Each measurement tool was successful at measuring
interface motion, but involved invasive methods of accessing the stem-cement interface,
in addition to being specific to loading mode type. For use in this thesis, however, a non-
invasive measuring tool capable of detecting motion under different loading conditions is
required, as future studies will be investigating implant micromotion under torsion
(Chapter 3 and 5), axial (Chapter 4 and 5), and bending (Chapter 6) loads. As such, the
purpose of this study was to develop and validate a measurement system capable of
detecting and measuring stem-cement interface micromotion during mechanical testing of
implant stems.
Optical measurement systems offer a non-invasive approach to measuring motion
in biomechanics studies. The choice of optical system depends on the requirements of
59
the measurement being made. For interface motion, which is expected to occur on the
order of micrometers, an appropriate optical system with high pixel resolution was
required. As such, the pilot GigE series camera with sensor size of 2454 x 2056 pixels;
3.45 μm/pixel, and telecentric lens was incorporated into the optical system. The
telecentric lens allowed for consistent image magnification with low distortion and
perspective errors. The source of illumination (i.e., axial diffuse illuminator) was also
specific to the needs of the measurement system, providing uniform light intensity with
low specular deflection. This ensured that there was no light scatter that may
compromise the resulting image, and influence the thresholding method used to detect the
markers.
From calibration of the optical system, it was determined that the resolution of the
system was appropriate for micrometer measurements, with pixel values on the order of
0.003 ± 0.002 mm (Appendix F).
The Bland and Altman plot was chosen to compare the agreement between the
optical system and the micrometer measurements, since it allowed comparison of the
differences between the measuring systems, based on the expected confidence interval
(i.e., limits of agreement) in which 95% of the differences were expected to fall (Bland
and Altman, 1999, 1986). Results from the Bland and Altman plot demonstrated good
agreement between both measuring tools, since the majority of scatter points fell within
95% confidence interval of the mean difference between the systems’ measurements
(Figure 2.5). However, for target vertical displacements of 0.5 mm, the optical system
consistently underestimated the measurement. This underestimate is believed to be due
to systematic error in obtaining large vertical displacements. In addition, as the target
displacements increased, the difference in the measurements between the optical system
and micrometer also increased as well, suggesting that the optical system was less
sensitive to measuring larger displacements from the micrometer. Since interface motion
is expected to be small (i.e., less than 200 µm) this observation was not of concern for the
intended application of the measurement system.
60
When comparing the error in the optical system based on known displacements
applied by the micrometer, it was found that for the range of measurements between
0.005 mm and 0.500 mm the percent error was +8.8% and -9.3% (Appendix E). This
accuracy measurement seems appropriate for the measuring capabilities required for
stem-cement interface motion. While there may be room for improvement, it was
determined that the error in the system was acceptable for its intended function of
comparing stem-cement interface motions.
With regards to the reliability of the optical system, ICC’s were used to compare
the correlation among repeated measures of the optical system. This statistic was
expected to give information on the precision of the optical system, and its reliability in
obtaining repeated measures with little variation. Results of the ICC demonstrated that
the correlation values for each of the image regions (i.e., Regions 1–9) were greater than
0.99 (Table 2.1). This agreement among the repeated measures illustrated that the optical
system was reliable at reporting ‘test-re-test’ measurements, which is important for multi-
trial testing.
The micrometer screw gauge was used as the gold-standard for assessing the
validity of the optical system, since it is a standard measurement tool with system
resolution on the order of 0.0001 inch (i.e., 0.0025 mm). However, because the optical
system was validated against another measurement tool, it is important to note that errors
in the measures obtained by the optical system may also be contributed to by errors
inherent in the micrometer screw gauge.
The color thresholding method of the optical system’s software detected markers
based on a single R-G-B value, which allowed distinction of the colored markers from
their surroundings. This was done by hand-tuning the R-G-B scale to a set value,
ensuring that only the makers appeared in the resultant thresholded image. While this
was a time consuming process to ensure a stable centroid was achieved, no concerns with
system performance were noted. However, a range of R-G-B values could have also
been used to capture the markers. It is possible that this range may have allowed easier
61
definition of the marker shapes and boundaries within the thresholded image, to achieve
the required stable centroid during static images.
The markers used for validation of the optical system were similar to those
expected to be used during mechanical testing in subsequent chapters, with regards to the
respective R-G-B values. However, because of the restrictions and difficulty of marker
attachment to the micrometer’s anvil and spindle, the markers were slightly larger in
shape compared to those expected to be used with mechanical testing. In addition, the
markers were positioned at the nearest proximity to the center of the anvil and spindle to
reduce offset motion during spindle rotation to the target displacement. However, offset
motion was still apparent, therefore analysis of relative marker motion in the horizontal
and vertical orientations were broken down into individual x and y components. Despite
meticulous methods to control for appropriate marker size and placement on the
micrometer’s anvil and spindle faces, the centroid tracking method of the optical system
may have been influenced by the non-uniformity of the markers. This may have further
contributed to the error detected in the optical measurement system.
2.5 CONCLUSION
Overall, this study was able to demonstrate the efficacy of the presented optical
system at meeting the system requirements needed for measuring interface motion.
Taking into consideration the reliability and accuracy of the system to measure
displacements on the order of micrometers, it was concluded that the optical system was
satisfactory for the application in this thesis, to compare interface motion of various stem
surface designs.
62
2.6 REFERENCES
Anglin, C., Wyss, U.P., 2000. Review of arm motion analyses. Proceedings of the
Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine
214, 541–555.
Bland, J., Altman, D., 1999. Measuring agreement in method comparison studies.
Statistical Methods in Medical Research 8, 135–160.
Bland, J.M., Altman, D.G., 1986. Statistical methods for assessing agreement between
two methods of clinical measurement. Lancet 1, 307–10.
Burke, D.W., O’Connor, D.O., Zalenski, E.B., Jasty, M., Harris, W.H., 1991.
Micromotion of cemented and uncemented femoral components. The Journal of
Bone and Joint Surgery; British Volume 73, 33–37.
Choi, D., Park, Y., Yoon, Y.-S.S., Masri, B.A., 2010. In vitro measurement of interface
micromotion and crack in cemented total hip arthroplasty systems with different
surface roughness. Clinical Biomechanics 25, 50–55.
Cristofolini, L., Teutonico, A.S., Monti, L., Cappello, A., Toni, A., 2003. Comparative in
vitro study on the long term performance of cemented hip stems: validation of a
protocol to discriminate between “good” and “bad” designs. Journal of
Biomechanics 36, 1603–1615.
Jamali, A.A., Lozynsky, A.J., Harris, W.H., 2006. The effect of surface finish and of
vertical ribs on the stability of a cemented femoral stem: an in vitro stair climbing
test. The Journal of Arthroplasty 21, 122–128.
Jeffers, J.R., Browne, M., Taylor, M., 2005. Damage accumulation, fatigue and creep
behaviour of vacuum mixed bone cement. Biomaterials 26, 5532–5541.
Lewis, G., 2011. Viscoelastic properties of injectable bone cements for orthopaedic
applications: state-of-the-art review. Journal of Biomedical Materials Research, Part
B, Applied Biomaterials 98, 171–191.
63
Maher, S.A., Prendergast, P.J., Lyons, C.G., 2001. Measurement of the migration of a
cemented hip prosthesis in an in vitro test. Clinical Biomechanics 16, 307–314.
Maletsky, L.P., Sun, J., Morton, N.A., 2007. Accuracy of an optical active-marker system
to track the relative motion of rigid bodies. Journal of Biomechanics 40, 682–685.
Mann, K.A., Miller, M.A., Verdonschot, N., Izant, T.H., Race, A., 2010. Functional
interface micromechanics of 11 en-bloc retrieved cemented femoral hip
replacements. Acta Orthopaedica 81, 308–317.
Myles, P.S., Cui, J., 2007. Using the Bland-Altman method to measure agreement with
repeated measures. British Journal of Anaesthesia 99, 309–311.
Race, A., Miller, M.A., Mann, K.A., 2010. Novel methods to study functional loading
micromechanics at the stem-cement and cement-bone interface in cemented femoral
hip replacements. Journal of Biomechanics 43, 788–791.
Shrout, P.E., Fleiss, J.L., 1979. Intraclass correlations: uses in assessing rater reliability.
Psychological Bulletin 86, 420–428.
Verdonschot, N., Huiskes, R., 1995. Dynamic creep behavior of acrylic bone cement.
Journal of Biomedical Materials Research 29, 575–581.
Zhang, H.Y., Brown, L., Barrans, S., Blunt, L., Jiang, X.Q., 2009. Investigation of
relative micromotion at the stem-cement interface in total hip replacement.
Proceedings of the Institution of Mechanical Engineers.Part H, Journal of
Engineering in Medicine 223, 955–964.
Zhou, H., Hu, H., 2008. Human motion tracking for rehabilitation—A survey.
Biomedical Signal Processing and Control 3, 1–18.
64
CHAPTER 3: THE EFFECT OF STEM SURFACE TREATMENT
ON THE TORSIONAL STABILITY OF TITANIUM AND COBALT
CHROME CEMENTED JOINT REPLACEMENT SYSTEMS
Overview: Stem surface treatment and substrate material are two design factors that can
contribute to the stability at the stem-cement interface, in resisting the effects of torsional
loads that occur at the upper limb joints. This chapter compared the torsional stability of
three stem surface treatments (smooth, beaded and plasma spray), applied to two
substrate materials (titanium and cobalt chrome), using the custom optical tracking
system described in Chapter 2 to measure interface rotation. 1
3.1 INTRODUCTION
Implant loosening is a complication that affects the success of joint replacement
systems, leading to revision surgery (Australian Orthopaedic Association, 2010; New
Zealand Orthopaedic Association, 2010). Loosening can be affected by implant stem
design (Verdonschot, 2005) and joint loading (Bergmann et al., 1995). For cemented
implants, a successful stem design has the capability of providing secure mechanical
fixation and stability at the stem-cement interface, resisting the effects of loosening
caused by various loading modes applied to the joint.
Stem surface treatment is one aspect of stem design that plays a role in cemented
implant stability. It has been demonstrated that cemented roughened surfaces
1 A version of this work is under review: Y.K. Hosein, G.J.W. King, C.E. Dunning (2013). “The Effect of
Stem Material and Surface Treatment on the Torsional Stability at the Metal-Cement Interface of Cemented
Upper Limb Joint Replacement Systems” Journal of Biomedical Materials Research: Part B Applied
Biomaterials.
65
experienced greater interface strength when compared to smooth polished surfaces (Chen
et al., 1998; Crowninshield, 1998; Walsh et al., 2004). These studies compared surface
treatment based on average roughness values (Ra), where surfaces used were classified as
polished (Ra < 1 µm), matte (Ra < 2 µm) or rough (Ra > 2 µm) (Verdonschot, 2005).
While this classification is relevant to micro surface modifications, it does not represent
surface treatments that are used clinically with cemented upper extremity stemmed
components (i.e., plasma spray and beaded). A study by Jeon et al. addressed the clinical
success of the Coonrad/Morrey Total Elbow implant which incorporated these treatments
(Jeon et al., 2012), however, no known in-vitro studies have specifically compared their
role in implant stability. Such information may be beneficial to the design of future joint
replacement systems.
Stem material is another aspect of stem design that may influence implant
stability. Titanium and cobalt chrome alloys are two conventional materials used with
implant stems (Buechel et al., 2012). With regards to cemented implants, however, there
has been much debate over the use of titanium stems with femoral components, due to
increased rates of loosening, as well as high sensitivity to corrosion (Hallam et al., 2004;
Schweizer et al., 2005). Despite lack of success with lower limb joint replacements,
cemented titanium implants remain extensively used in upper limb applications (Van der
Lugt and Rozing, 2004). In such cases, the type of stem material may influence the
success of surface treatments applied to the implant stem, by affecting the strength of the
bond formed between the stem and coating during the finishing process (Davis, 2003).
Joint loading can also affect implant loosening. Torsional loading that occurs at
the joint is one such example that can lead to failure of the prosthesis. This failure
mechanism can be initiated during lifting movements with the elbow in a flexed position.
During this motion, internal rotational loads are directed at the humerus of the elbow
joint (Van der Lugt et al., 2010). Such is the case in lifting with an abducted arm, or the
throwing motion of the arm.
Considering the role stem surface roughness plays in implant stability, as well as
the lack of literature comparing surface treatments that are clinically relevant to cemented
66
upper extremity stemmed components, the purpose of this study was to investigate the
effect of stem surface treatment on the torsional stability of both titanium and cobalt
chrome cemented stems in-vitro.
3.2 MATERIALS AND METHODS
Thirty generic implant stems of circular cross-section (Ø = 8 mm) were custom-
machined from cobalt chrome (n = 15) and titanium (n = 15) by Tornier S.A.S.
(Montbonnot Saint Martin, France) (See Appendix G for engineering drawings). Each
stem was given a polished smooth surface. The fifteen stems of each material were split
into three equal groups. One group of stems retained their smooth surface (n = 5), while
the remaining groups had two different surface treatments: beaded (n = 5) and plasma
sprayed (n = 5), applied over a 20 mm stem length region by Orchid Bio-Coat
(Southfield, MI, USA) (Figure 3.1). The plasma sprayed surfaces were titanium plasma
spray (TPS) and had a Ra value of 48.5 ± 3.9 µm. The beaded surfaces consisted of one
layer of beads, with a bead diameter of approximately 500 µm.
All stems were centralized in square aluminum tubes, and potted using vacuum-
mixed PMMA bone cement (Simplex P
, Stryker
, Kalamazoo, MI, USA) to a fixed
depth of 20 mm, such that there was full coverage of the surface treated regions (for
detailed potting and centralization techniques, refer to Appendix H). The stems were
maintained in air at 22 ºC for 24 hours during curing. Subsequently, the potted stems
were secured in a materials testing machine (Figure 3.2) (Instron
8874, Norwood, MA,
USA) and cyclically tested at 1.5 Hz under torsion. Custom machined fixturing ensured
that stems were fully centralized during testing (Appendix G.10), allowing the
application of pure torsional loads with minimal contribution of off-axis loading, as
monitored by the 6 degrees of freedom load cell of the materials testing machine.
Loading cycled from a lower limit of 0 Nm to an upper limit that started at 1 Nm,
increasing in increments of 1 Nm every 100 cycles to a maximum of 30 Nm, or until a
67
Figure 3.1: Surface Treated Stems used for Torsional Testing
(A) Smooth surface, (B) beaded treatment, and (C) plasma spray treatment. All stems
were cemented to a fixed depth of 20 mm, to ensure coverage of only the surface treated
region of the stem.
68
rapid increase in rotation of the stem occurred without resistance; the latter being termed
catastrophic failure.
Motion at the stem-cement interface was quantified using the custom optical
tracking system described in Chapter 2 of this thesis (Basler Pilot GigE Camera [Basler,
Ahresnburg, Germany]; Opto Engineering Telecentric Lens [Opto Engineering, Mantua,
Italy]; Axial Diffuse Illuminator [Advanced Illumination, Rochester, VT, USA]; and
LabVIEW Vision Acquisition System [National Instruments,Austin, TX, USA]). This
system incorporated a colour thresholding method to optically track the centroid of
markers placed on the stem and cement (Figure 3.2), to determine their relative distances
throughout loading.
Stem-cement interface strength and interface stability were both quantified.
Interface strength was determined from the values of torque at failure, as obtained from
the materials testing machine. Interface stability was determined from the relative
rotational displacement between the markers on the stem and cement immediately prior
to failure, as obtained from the optical tracking system (i.e., termed “interface toggle”;
Figure 3.3).
Two-way analyses of variance with post-hoc Student-Newman-Keuls tests
(α = 0.05) were used to examine the role of stem surface treatment and stem material on
torsional stability, based on the measures of interface strength and stability.
3.3 RESULTS
Overall, torsional load targets were achieved with minimal off-axis loads and
moments detected. To assess stem-cement interface toggle, graphs of relative stem
rotation versus the number of loading cycles were created (Figure 3.3). In addition to
interface toggle, which was measured just prior to failure, the presence of offset stem
motion was observed in these graphs. This motion, which was representative of the
69
Figure 3.2: Experimental Set-up for Torsional Testing of Implant Stems
Cemented stem within the aluminum tube, secured in an Instron® Materials Testing
Machine. Shown in the inset are optical markers attached to the stem and the cement.
The stem was cyclically loaded under torque, and a camera system was used to detect
stem-cement interface motion throughout loading.
70
Figure 3.3: Stem Rotation for Titanium and Cobalt Chrome Surface Treated Stems
The graphs of relative stem rotation display interface toggle for all stems, which was a
measure of the stem-cement interface stability just prior to failure. Offset stem motion,
as represented by the deviation of the titanium beaded and titanium plasma spray stems
away from their points of origin, is also shown. Stems that did not complete the full 3000
cycles of the loading protocol experienced catastrophic failure, as depicted by the rapid
increase in relative stem rotation.
71
deviation of the stem away from its point of origin, qualitatively appeared to vary with
both surface treatment and substrate material, but was not quantitatively analyzed.
From the initial results, it was noted that stem rotation increased with increasing
applied torque (Figure 3.3). Therefore, due to the staircase loading protocol employed,
the interface toggle was normalized to the value of torque at failure.
Of the three stem surfaces tested, the smooth stem was the only surface that
consistently experienced catastrophic failure for both titanium and cobalt chrome stems.
The beaded surface experienced catastrophic failure with the cobalt chrome stem material
only, and post-testing inspection found that this resulted in mechanical damage to the
beaded coating (Figure 3.4). The plasma spray stems, with both the titanium and cobalt
chrome stem materials, never experienced catastrophic failure (Figure 3.5).
As such, there was an overall effect of surface treatment on interface strength
(p < 0.001), where smooth stems required less torque to cause interface failure than either
the beaded (p < 0.05) or plasma spray stems (p < 0.05), and plasma spray was also
superior to beaded (p < 0.001) (Figure 3.5). Stem material also had an overall effect on
strength (p = 0.001), where titanium stems experienced higher torques at failure than
cobalt chrome stems. In addition, the ANOVA demonstrated a significant interaction
between stem surface treatment and stem material (p < 0.001). Thus, appropriate one-
way ANOVAs were conducted, and differed from the main effect in that there was no
difference between material for plasma spray only (p = 1.0), and no difference between
beaded and plasma spray for titanium stems only (p = 1.0).
With regards to interface stability, stem surface treatment had an overall effect
(p = 0.002) with smooth stems, demonstrating more interface toggle than either beaded
(p < 0.05) and plasma spray (p < 0.05) treatments (Figure 3.6). However, the effect of
surface treatment depended on stem material (i.e., significant interaction; p = 0.001).
One-way ANOVA’s showed that all three surfaces were different from one another for
titanium stems (p < 0.05), with a stability ranking of: beaded > plasma spray > smooth.
By comparison, surface treatment did not have an effect on stability for cobalt chrome
72
Figure 3.4: Inspection of Stem Surface Treatments Post-Torsion Tests
(A) Smooth titanium (left) and smooth cobalt chrome (right), (B) beaded titanium (left)
and beaded cobalt chrome (right), and (C) plasma spray titanium (left) and plasma spray
cobalt chrome (right). The beaded cobalt chrome was the only stem that experienced
mechanical damage to its coating at failure, as depicted by (**) in the figure above.
73
Figure 3.5: Torque at Failure for Surface Treated Stems
Graph showing torque at failure (i.e., as a measure of interface strength) for smooth,
beaded and plasma sprayed surface treatments, on titanium and cobalt chrome stems.
The bars marked with white stars represent those stems that consistently experienced
catastrophic failure.
74
Figure 3.6: Normalized Interface Toggle for Surface Treated Stems in Torsion
Graph showing normalized interface toggle prior to failure (i.e., as a measure of interface
motion) for smooth, beaded and plasma sprayed surface treatments, on titanium and
cobalt chrome stems. The bars marked with white stars represent those stems that
consistently experienced catastrophic failure.
75
stems (p > 0.05). Stem material also showed an overall effect on interface stability,
where titanium stems experienced more toggle than cobalt chrome stems (p = 0.031).
3.4 DISCUSSION
Improved implant longevity remain a continuous goal for orthopaedic implant
research. Therefore, investigating factors affecting implant loosening is important for
understanding possible failure mechanisms. Stem design at the stem-cement interface is
one factor that may contribute to the overall stability of the implant system, and
subsequent loosening (Barrack, 2000).
Stem surface treatments have been incorporated into implant designs to improve
implant fixation. Surface treatments such as sintered beads and plasma spray are
typically incorporated into non-cemented implant designs, with the expectation that the
porous surface would allow for bony-in-growth onto the implant stem. However, these
surface treatments are also utilized with cemented upper limb implant designs, such as
the Zimmer®
Coonrad/Morrey Total Elbow (Zimmer, Warsaw, IN, USA) (titanium
plasma spray and beads on titanium stems), Latitude® EV Total Elbow Prosthesis
(Tornier SAS, Montbonnot Saint Martin, France) (titanium plasma spray on cobalt
chrome stems), and Discovery® Elbow System (Biomet, Warsaw, IN, USA) (titanium
plasma spray on titanium, or cobalt chrome stems). It is expected that the viscous bone
cement fills the spaces of the porous stem surface during cement application, and when
cured, initiates mechanical interlock of the bulk cement to the stem surface. This
mechanical interlock can improve implant fixation by allowing for better anchoring of the
implant stem within the cement. In addition, during the lifespan of the implant system,
these surface treatments are likely to provide frictional resistance to implant motion that
may be caused as a result of loading occurring at the joint.
Cobalt chrome and titanium are two established materials among the orthopaedic
community. Within recent years, titanium has been a controversial material for use in
femoral cemented stem designs, since it has been reported to experience increased
76
loosening rates, as well as high sensitivity to corrosion (Jergesen and Karlen, 2002).
Despite this, titanium is still commonly used in cemented upper-limb implant
applications (Sanchez-Sotelo, 2011; van der Lugt and Rozing, 2004). As such, this study
aimed to examine the effect of stem surface treatment on the stability of cemented
titanium and cobalt chrome implant stems under torsional loading.
Stem surface treatment affected stem stability, where beaded and plasma sprayed
surfaces experienced more resistance to rotation prior to failure when compared to the
smooth stems. In addition, the smooth stems consistently experienced catastrophic
failure at the lowest torque values for both the titanium and cobalt chrome stems.
Beaded stem surfaces appeared to perform better with a titanium stem material, as
observed from their resistance to catastrophic failure, when compared to cobalt chrome
stems. Post-testing inspection of the stems showed that catastrophic failure occurred as a
result of the beaded coating being completely ripped off the stem surface (Figure 3.4).
Thus, it is likely that the torque experienced at failure is more representative of the
bonding strength of the beaded coating to the cobalt chrome stem, which is a factor of the
coating fabrication process, and not necessarily a representation of the stem-cement
interface strength.
For titanium stems examined in our study, the beaded stems experienced less
interface toggle than the plasma spray stems, with no difference seen in interface strength
(Figure 3.6 and 3.5, respectively). This may be explained based on the mechanism
explained previously, involving mechanical interlock of the stem surface with cement.
Beaded coatings have greater surface deviation based on the diameter of the sintered
beads. This may allow for deeper infiltration of bone cement, subsequently improving
the anchoring and stability of the implant system.
Overall results showed that titanium stems experienced more interface toggle
throughout loading than cobalt chrome stems (Figure 3.3). This seems reasonable taking
into consideration their respective material properties. Titanium is more ductile
compared to cobalt chrome, as defined by its lower elastic modulus (Navarro et al.,
77
2008), and therefore likely to result in greater stem twisting with respect to the surface of
the cement.
From Figure 3.3 it was observed that in addition to interface toggle, there were
also offsets of the relative stem rotation for stems that did not experience catastrophic
failure (i.e, plasma sprayed and beaded titanium). The offsets represent deviation of the
stem away from the point of origin with increased loading. This deviation may be due to
cement creep at the interface as a result of cement exposure to fatigue loading (Jeffers et
al., 2005; Lewis, 2011, 1997). One method of defining implant loosening is detection of
implant migration within the host bone. As such, it is important to acknowledge bone
cement creep when defining interface stability as a measure of implant motion only, since
cement creep may cause stem motion without interface loosening.
Plasma spray coatings are applied using a thermal spray in a reduced pressure
inert gas chamber at high temperatures, to produce an irregular surface on the stem. In
comparison, beaded coatings are applied via a sintering process, where the beads are
bonded to the substrate surface of the stem (Ryan et al., 2006). The strength of the bond
formed is affected by the carbon content of the base metal substrate, where a lower
carbon content reduces the effectiveness of the sintering process, causing an inferior bond
at the level of the substrate surface and beads (Davis, 2003). Thus, it is possible that the
effect of substrate carbon content on bonding strength may have played a role in our
study, where post-testing inspection of the cobalt chrome stems found the beaded coating
completely ripped off the stem surfaces during failure, but remained intact for the
titanium stems (Figure 3.4).
For this study, only the torsional stability of various surface treatments was
investigated. The loading protocol used was chosen based on a review of available
literature directly measuring the torque through the use of instrumented implants at
various joints (i.e., shoulder, hip, knee) (Bergmann et al., 2007, 2001; Kutzner et al.,
2010), as well as incorporation of a cyclic staircase method at a rate of 1.5 Hz to mimic
fatigue loading that typically occurs in-vivo. This is in keeping with the literature, for
example, femoral components have been tested using loading rates between 1–3 Hz (Bell
78
et al., 2007; Britton et al., 2004; Hernández-Rodríguez et al., 2005; Westphal et al.,
2006; Wilson et al., 2009). This protocol ensured comparison of the surface treatments
based on clinically relevant joint loads, since upper limb joints experience similar loading
patterns to those at the lower limbs (Goldberg et al., 1988). Anatomically, however, the
joint is exposed to a variety of loading modes in addition to torque, which typically work
in a combined state at the joint. Therefore, future studies will look at investigating stem
surface treatment on implant stability under other loading modes.
The optical tracking system used in our study calculated the relative distances
between optical markers attached to the exposed surfaces of the stem and cement, to give
measurements of interface motion throughout loading. This method is a better
representation of interface motion compared to marker attachment to the implant only,
since it measures motion solely at the stem-cement interface, regardless of any possible
motion in the surrounding fixtures. The measurement, however, does not discern
interface motion along the length of the cemented stem. Therefore, future studies will
incorporate linear variable differential transducers embedded in the cement mantle, or
imaging techniques to detect internal motion of the stem and cement.
Square aluminum tubes were used instead of cadaveric bone for cementing of the
implant stems. Although this does not represent a clinically cemented intramedullary
component, aluminum tubes were chosen since the study was only interested in the
effects at the stem-cement interface. The use of aluminum tubes did have the benefit of
providing a controlled cementing and testing environment, reducing any variation in the
results that may be caused by varying bone quality.
With regards to the implant stem designs, circular cross-section generic stems,
with 20 mm length surface treated regions were used to compare implant stability. This
design is not representative of a specific joint implant stem in terms of its geometry and
surface treated area, but allowed for comparison of implant stability solely dependent of
surface treatment. Circular cross-section implant stems provided the least resistance to
torque (Kedgley et al., 2007), and as such, it was intentionally incorporated into the study
design to minimize the effects of stem geometry on implant stability. Similarly, a 20 mm
79
length surface treatment region was used, since pilot testing showed that greater surface
coverage did not allow for interface failure in the desired testing period, and reduced
surface coverage was not optimal for relative comparison of surface treatments.
The findings from this study suggest that stem surface treatment influences
implant stability, offering greater interface strength and resistance to motion than a
smooth surface. When comparing surface treatments, the plasma spray finish offered
superior interface strength compared to the beaded finish with cobalt chrome stems;
however there was no difference in their performance with titanium stems. With regards
to interface stability, the beaded finish was more stable than the plasma spray for titanium
stems, but showed no difference for cobalt chrome stems. In addition to surface
treatment, our study demonstrated that stem material affected implant stability, with
titanium stems experiencing greater interface strength but reduced resistance to interface
motion than cobalt chrome stems.
3.5 CONCLUSION
For surface treated stems tested under cyclic torsional loading, plasma spray
performed better than beaded treatments on a cobalt chrome stem substrate, and beaded
treatments performed comparable to plasma spray treatments on a titanium stem
substrate.
80
3.6 REFERENCES
Australian Orthopaedic Association, T., 2010. National Joint Registry, Hip and Knee
Arthroplasty Annual Report 2010,
http://www.dmac.adelaide.edu.au/aoanjrr/documents/aoanjrrreport_2010.pdf.
Barrack, R.L., 2000. Early failure of modern cemented stems. The Journal of
Arthroplasty 15, 1036–1050.
Bell, C.G., Weinrauch, P., Pearcy, M., Crawford, R., 2007. In vitro analysis of exeter
stem torsional stability. The Journal of Arthroplasty 22, 1024–1030.
Bergmann, G., Deuretzbacher, G., Heller, M., Graichen, F., Rohlmann, A., Strauss, J.,
Duda, G.N., 2001. Hip contact forces and gait patterns from routine activities.
Journal of Biomechanics 34, 859–871.
Bergmann, G., Graichen, F., Bender, A., Kääb, M., Rohlmann, A., Westerhoff, P., 2007.
In vivo glenohumeral contact forces--measurements in the first patient 7 months
postoperatively. Journal of Biomechanics 40, 2139–2149.
Bergmann, G., Graichen, F., Rohlmann, A., 1995. Is staircase walking a risk for the
fixation of hip implants? Journal of Biomechanics 28, 535–553.
Britton, J.R., Lyons, C.G., Prendergast, P.J., 2004. Measurement of the Relative Motion
Between an Implant and Bone under Cyclic Loading. Strain 40, 193–202.
Buechel, F.F., Pappas, M.J., SpringerLink, 2012. Properties of Materials in Orthopaedic
Implant Systems. In: Principles of Human Joint Replacement. Springer Berlin
Heidelberg, Berlin, Heidelberg, pp. 1–35.
Chen, P.C., Pinto, J.G., Mead, E.H., D’Lima, D.D., Colwell Jr., C.W., 1998. Fatigue
model to characterize cement-metal interface in dynamic shear. Clinical
Orthopaedics and Related Research 350, 229–236.
81
Crowninshield, R., 1998. Cemented femoral component surface finish mechanics.
Clinical Orthopaedics and Related Research 90–102.
Davis, J.R., 2003. Coatings. In: Handbook of Materials for Medical Devices. ASM
International, Materials Park, OH, pp. 179–194.
Goldberg, V.M., Figgie 3rd, H.E., Inglis, A.E., Figgie, M.P., 1988. Total elbow
arthroplasty. The Journal of Bone and Joint Surgery; American Volume 70, 778–
783.
Hallam, P., Haddad, F., Cobb, J., 2004. Pain in the well-fixed, aseptic titanium hip
replacement. The role of corrosion. The Journal of Bone and Joint Surgery; British
Volume 86, 27–30.
Hernández-Rodríguez, M. a. L., Mercado-Solís, R.D., Pérez-Unzueta, a. J., Martinez-
Delgado, D.I., Cantú-Sifuentes, M., 2005. Wear of cast metal–metal pairs for total
replacement hip prostheses. Wear 259, 958–963.
Jeffers, J.R., Browne, M., Taylor, M., 2005. Damage accumulation, fatigue and creep
behaviour of vacuum mixed bone cement. Biomaterials 26, 5532–5541.
Jeon, I.H., Morrey, B.F., Sanchez-Sotelo, J., 2012. Ulnar component surface finish
influenced the outcome of primary Coonrad-Morrey total elbow arthroplasty.
Journal of Shoulder and Elbow Surgery 21, 1229–1235.
Jergesen, H.E., Karlen, J.W., 2002. Clinical outcome in total hip arthroplasty using a
cemented titanium femoral prosthesis. The Journal of Arthroplasty 17, 592–599.
Kedgley, A.E., Takaki, S.E., Lang, P., Dunning, C.E., 2007. The effect of cross-sectional
stem shape on the torsional stability of cemented implant components. Journal of
Biomechanical Engineering 129, 310–314.
Kutzner, I., Heinlein, B., Graichen, F., Bender, A., Rohlmann, A., Halder, A., Beier, A.,
Bergmann, G., 2010. Loading of the knee joint during activities of daily living
measured in vivo in five subjects. Journal of Biomechanics 43, 2164–2173.
82
Lewis, G., 1997. Properties of acrylic bone cement: state of the art review. Journal of
Biomedical Materials Research 38, 155–182.
Lewis, G., 2011. Viscoelastic properties of injectable bone cements for orthopaedic
applications: state-of-the-art review. Journal of Biomedical Materials Research Part
B, Applied Biomaterials 98, 171–191.
Navarro, M., Michiardi, A., Castano, O., Planell, J.A., Castaño, O., 2008. Biomaterials in
orthopaedics. Journal of the Royal Society, Interface 5, 1137–1158.
New Zealand Orthopaedic Association, T.N.Z.J.R., 2010. Eleven Year Report- January
1999 to December 2009., http://www.cdhb.govt.nz/njr/reports/A2D65CA3.pdf.
Ryan, G., Pandit, A., Apatsidis, D.P., 2006. Fabrication methods of porous metals for use
in orthopaedic applications. Biomaterials 27, 2651–2670.
Sanchez-Sotelo, J., 2011. Total elbow arthroplasty. The Open Orthopaedics Journal 5,
115–123.
Schweizer, A., Luem, M., Riede, U., Lindenlaub, P., Ochsner, P.E., 2005. Five-year
results of two cemented hip stem models each made of two different alloys.
Archives of Orthopaedic and Trauma Surgery 125, 80–86.
Van der Lugt, J.C., Rozing, P.M., 2004. Systematic review of primary total elbow
prostheses used for the rheumatoid elbow. Clinical Rheumatology 23, 291–298.
Van der Lugt, J.C., Suarez, D.R., van der Steenhoven, T.J., Nelissen, R.G., 2010. Minor
influence of humeral component size on torsional stiffness of the Souter-Strathclyde
total elbow prosthesis. International Orthopaedics 34, 1213.
Verdonschot, N., 2005. Philosophies of stem designs in cemented total hip replacement.
Orthopedics 28, s833–840.
Walsh, W.R., Svehla, M.J., Russell, J., Saito, M., Nakashima, T., Gillies, R.M., Bruce,
W., Hori, R., 2004. Cemented fixation with PMMA or Bis-GMA resin
83
hydroxyapatite cement: effect of implant surface roughness. Biomaterials 25, 4929–
4934.
Westphal, F.M., Bishop, N., Honl, M., Hille, E., Puschel, K., Morlock, M.M., 2006.
Migration and cyclic motion of a new short-stemmed hip prosthesis--a
biomechanical in vitro study. Clinical Biomechanics 21, 834–840.
Wilson, L.J., Bell, C.G.R., Weinrauch, P., Crawford, R., 2009. In vitro cyclic testing of
the Exeter stem after cement within cement revision. The Journal of arthroplasty 24,
789–94.
84
CHAPTER 4: THE EFFECT OF STEM SURFACE TREATMENT
AND MATERIAL ON PISTONING OF ULNAR COMPONENTS IN
LINKED CEMENTED ELBOW PROSTHESES
Overview: The ulnar component of a total elbow replacement can fail via stem
“pistoning”, which occurs as a result of axial loads generated at the ulnohumeral joint in
routine daily activities. The application of stem surface treatments to titanium and cobalt
chrome elbow systems has resulted in improved clinical success, but with varied
responses. This chapter compares the stability response of smooth, beaded and plasma
spray surface treatments, applied to titanium and cobalt chrome stems, in resisting axial
forces at the stem-cement interface. Results of this study explain the relationship between
stem surface treatments and substrate material in overall implant stability, as well as
suggest various failure mechanisms associated with each of the surface treatments. 2
4.1 INTRODUCTION
Ulnar component pistoning has been described as one of the main failure
mechanisms of total elbow prostheses (Cheung and O’Driscoll, 2007), leading to
loosening of the implant system. Joint reaction forces that result in tension/compression
loads acting across the ulnohumeral joint occur when the elbow is in the flexed position
and can lead to this pistoning effect (Amis, 2012; Goldberg et al., 1988; Johnson and
King, 2005). Stem design factors, including the application of surface treatments, are
expected to resist the effects of mechanical loosening caused by these forces (Evans et
2 A version of this work has been published: Y.K. Hosein, G.J.W. King, C.E. Dunning (2013). “The Effect
of Stem Surface Treatment and Material on Pistoning observed in Ulnar Components of Linked Cemented
Elbow Prostheses” Journal of Shoulder and Elbow Surgery. [E-pub ahead of print; PMID: 23668920] (See
Appendix I for the letter of permission)
85
al., 1988). With surface treatments, improved cement fixation can be hypothesized based
on the premise of a mechanical interlock formed between the bulk cement and treated
stem surface. For ulnar components specifically, a clinical study by Jeon et al., showed
that various surface treated stems experienced different rates of loosening (Jeon et al.,
2012); however, no known in-vitro studies have compared the success of these surface
treatments relative to one another.
Sintered beads and thermal plasma sprays are two common surface treatments
used in ulnar stem designs. The Coonrad/Morrey Total Elbow (Zimmer Inc.), in
particular, has modified its titanium prosthesis design over the years based in part on
varying these two surface treatments (Jeon et al., 2012). In addition, the Latitude EV
(Tornier Inc.) (cobalt chrome stem) and Discovery Elbow System (Biomet Inc.) (titanium
stem) incorporate a plasma spray surface treatment in their ulnar component designs.
When taking into consideration the varying surface topographies of the respective surface
treatments (i.e., beaded and plasma spray), it can be hypothesized that the type of stem
surface treatment may affect the strength of the mechanical interlock formed between the
stem and cement. Comparisons of this mechanical interlock can offer insight into failure
mechanisms associated with these surface treated implants.
Titanium and cobalt chrome alloys are both used as the substrate material in
elbow prosthesis stem designs. Stem substrate material may play a role in the success of
the applied surface treatment, since the composition of the substrate can influence the
strength of stem-treatment bond formed during treatment process (Davis, 2003). As
such, it is important to consider stem material when investigating the role of surface
treatments in prosthesis loosening.
Therefore, the purpose of this in-vitro study was to investigate the role of stem
surface treatment and substrate material on the stability of a simulated ulnar implant stem
subjected to cyclic compression loading.
86
4.2 MATERIALS AND METHODS
Sixty smooth, circular, implant stems (Ø = 8 mm) were custom machined from
both titanium (n = 30), and cobalt chrome (n = 30) by Tornier S.A.S. (Grenoble, France).
For each material, the stems were sub-divided into five equal groups. The first group
retained their full length smooth surface (n = 6), while the remaining four groups had
standard commercially employed stem surface treatments applied by Orchid Bio-Coat
(Southfield, MI, USA). Two groups received a beaded surface treatment with coverage
lengths of 20 mm (n = 6) and 10 mm (n = 6), while two groups received a plasma spray
surface treatment with coverage lengths of 20 mm (n = 6) and 10 mm (n = 6) (Figure 4.1)
(Appendix G). The beaded surfaces consisted of one layer of beads, with a bead diameter
of approximately 500 µm. The plasma sprayed surfaces were titanium plasma spray
(TPS) and had a Ra value of 48.5 ± 3.9 µm.
All stems were potted to a fixed depth of 20 mm in square aluminum tubes using
vacuum-mixed PMMA bone cement (Simplex P
, Stryker
, Kalamazoo, MI, USA), such
that there was full coverage of the surface treated regions. For the 10 mm surface treated
stems, this potting method allowed for cement coverage of the 10 mm treated region, as
well as an additional 10 mm of proximal smooth stem surface (Figure 4.1). The potted
stems were maintained in air at 22 ºC for 24 hours during curing. Subsequently, they
were secured in a materials testing machine (Figure 4.2) (Instron
8874, Norwood, MA,
USA). Custom machined fixturing ensured that stems were fully centralized during
testing (Appendix G.10), accommodating stem push-out under compression, with
minimal contribution of off-axis loading, as monitored by the 6 degrees of freedom load
cell of the materials testing machine. A Delrin® stopper was placed inside the aluminum
tube at the base of the cement mantle to ensure push-out of the stem only, without
slipping of the cement mantle within the aluminum tube.
Loading cycled at 1.5 Hz under compression, keeping loads between 500 N and
an upper limit. The upper limit started at 1000 N and increased in increments of 1000 N
every 100 cycles to a maximum 10000 N, after which it cycled for a further 25000 cycles
87
Figure 4.1: Surface Treated Implant Stems used for Compression Testing
Implant stems with various surface treatments: (A) smooth, (B) 20 mm length beaded
treatment, (C) 10 mm length beaded treatment, (D) 20 mm length plasma spray
treatment, and (E) 10 mm length plasma spray treatment. Stems were cemented to a
fixed 20 mm depth, as highlighted by the region between the parallel lines. For the
10 mm length surface treated stems, this allowed full cement coverage of the surface
treated region, as well as an additional 10 mm length of proximal smooth stem surface.
88
to complete the testing protocol (i.e., total 25900 cycles). Failure of the stem-cement
interface was defined as 2 mm push-out of the stem relative to the cement, termed
‘catastrophic failure’, or until completion of the loading protocol.
Motion at the stem-cement interface was quantified using the custom optical
tracking system described in Chapter 2 (Basler Pilot GigE Camera, Ahresnburg,
Germany; Opto Engineering Telecentric Lens, Mantova, Italy; and LabVIEW Vision
Acquisition System, National Instruments, Austin, TX, USA). This system incorporated
a colour thresholding method to optically track the centroid of markers placed on the
stem and cement (Figure 4.2), to determine their relative distances throughout loading.
Immediately following testing, the cemented stems were placed into Acetone
(Caledon Laboratories Ltd., Georgetown, ON, Canada) for 24 hours to allow dissolution
of the surrounding bone cement. The stems were subsequently cleaned and visually
inspected to determine the presence of any surface treatment damage associated with
testing.
Stem-cement interface strength and stem motion were both quantified. Interface
strength was determined from the number of cycles required to cause failure for each of
the stems, as obtained from the materials testing machine. Stem motion was determined
from the relative distances between the markers on the stem and cement prior to failure
(i.e., termed “interface toggle”) (Figure 4.2). For those stems that reached 10000 N in the
loading protocol, and survived beyond the first 100 cycles, an additional measure of stem
motion (i.e., termed “global motion”) was used. Global motion measured the
displacement of the stem within the camera’s coordinate system (i.e., not relative to the
cement) from 901 cycles (i.e., start of the 10000 N load step) until failure, and allowed
relative comparison of surface treatments at the constant 10000 N load level (Figure 4.2).
Two-way analyses of variance with post-hoc Student-Newman-Keuls tests
(α = 0.05) were used to examine the role of stem surface treatment and substrate material
on cemented stem stability, based on the measures of interface strength and stem motion.
89
Figure 4.2: Experimental Set-up for Compression Testing of Implant Stems
Schematic of the cemented stem in the materials testing machine, showing application of
a compressive load. The camera was used to track markers placed on the stem and
cement throughout loading as shown in the inset image. The relative distance between
the stem and cement was used to determine interface toggle (I.T.) just prior to failure.
The change in global motion of stems from 901 cycles (i.e., start of 10000 N load step)
until failure (G.S.M) was used to compare axial motion of surface treated stems at a
constant load level.
90
4.3 RESULTS
Post-testing visual inspection of the stems found that the 10 mm beaded titanium
and 10 mm beaded cobalt chrome stems were the only surfaces to experience mechanical
damage at failure (Figure 4.3). This was observed as debonding of the beaded treatments
from the stem surfaces. All other stem surfaces remained intact, with no visual damage
evident.
Survival curves illustrated that stem surface treatment did affect stem stability
under compression loading, where the 20 mm length, beaded stems outlasted the other
stem surfaces (Figure 4.4). Data obtained from the optical tracking system showed that
stem motion increased simultaneously with the cyclic staircase loading protocol (Figure
4.5), and as such, all motion data (i.e., interface toggle and global stem motion) was
normalized to their respective loads and cycles, to allow for relative comparison of stem
surfaces.
With regards to cycles to failure, the two-way ANOVA found an overall effect of
surface treatment (p < 0.05) and no overall effect of substrate material (p = 0.25), but
there was a significant interaction between these factors (p = 0.02) (Figure 4.6A).
Therefore, one-way ANOVAs were performed, and showed that for titanium, the 20 mm
beaded stems outlasted all other treatments (p < 0.05). For cobalt chrome, the 20 mm
beaded stems outlasted all other treatments (p < 0.05), but the 20 mm plasma spray stem
also performed better than the 10 mm beaded, 10 mm plasma spray, and smooth stems
(p < 0.05).
With regards to interface toggle, the two-way ANOVA again found an overall
effect of surface treatment (p < 0.05), no overall effect of substrate material (p = 0.78),
and a significant interaction between these factors (p = 0.01) (Figure 4.6B). One-way
ANOVAs showed that for titanium, the 20 mm beaded, 10 mm beaded, and 20 mm
plasma spray stems experienced less toggle than the 10 mm plasma spray and smooth
(p < 0.05), with the 10 mm plasma spray stem exhibiting less toggle than the smooth
(p < 0.05).
91
Figure 4.3: Inspection of Stem Surface Treatments Post-Compression Tests
(A) Titanium (left) and cobalt chrome (right) 20 mm length beaded stems, (B) titanium
(left) and chrome (right) 10 mm length beaded stems, (C) titanium (left) and cobalt
chrome (right) 20 mm length plasma spray stems, and (D) titanium (left) and cobalt
chrome (right) 10 mm length plasma spray stems. The 10 mm length beaded titanium
and cobalt chrome stems were the only stems to experience debonding of the treatment
from the stem surface at failure, as highlighted by the starred region.
92
Figure 4.4: Survival Curves for Titanium and Cobalt Chrome Stems
Survival curves for titanium (top), and cobalt chrome (bottom) stems. The 20 mm length
beaded stem experienced longest survival, while smooth stem surface experienced
shortest survival, as defined by the number of cycles required to cause failure.
93
Figure 4.5: Stem Motion for Titanium and Cobalt Chrome Stems under
Compression
Representative graphs of stem motion relative to cement (top row), and global stem
motion (bottom row) for titanium and cobalt chrome stems. The 20 mm length beaded
stem did not experience catastrophic failure, as defined by 2 mm of stem push-out, for
eleven of the twelve stems tested.
94
Figure 4.6: Interface Stability Offered by the Surface Treated Stems in
Compression
(A) Cycles at failure: For titanium, the 20 mm beaded treatment experienced the most
cycles compared to all stem surfaces (p < 0.05¥). For cobalt chrome, the 20 mm beaded
stem outlasted the other treatments (p < 0.05*), with 20 mm length plasma spray
treatment also showing greater cycles to failure than the other stem surfaces (p < 0.05†).
(B) Interface Toggle prior to failure: For titanium, the 20 mm beaded, 10 mm beaded, and
20 mm plasma spray stems showed the smallest magnitudes of interface toggle
(p < 0.05‡), with the 10 mm plasma spray stem experiencing less toggle than the smooth
(p < 0.05§). For cobalt chrome, the 20 mm beaded, and 20 mm plasma spray treatments
showed reduced interface toggle compared to all other stems (p < 0.05¤).
95
For cobalt chrome, the 20 mm beaded and 20 mm plasma spray stems experienced less
toggle than the 10 mm length surface treatments and smooth stems (p < 0.05).
When comparing the global motion of surface treated stems at a constant 10000 N
load level, it was found that surface treated stems demonstrated various magnitudes of
stem motion (Figure 4.7). Overall, the 20 mm and 10 mm beaded treatments experienced
the least stem motion (p < 0.05), with the 20 mm plasma spray treatment also showing
less stem motion than the 10 mm plasma spray treatment (p < 0.05). Cobalt chrome
stems showed less motion than titanium (p = 0.02).
4.4 DISCUSSION
Surface treatments have been added to cemented ulnar components of total elbow
arthroplasties with the aim of improving stem fixation and long term stability (Jeon et al.,
2012). Bone cement is considered a connecting agent, much like a grout material. When
used with a surface treated component, it is expected to fill the spaces of the treated
surface, initiating a mechanical interlock between the stem and cement. The strength of
the interlock formed may subsequently improve immediate and long term stem fixation.
The evolution of ulnar component designs over the last decade has incorporated
different stem surface treatments, specifically sintered beads and plasma spray, with the
goal of improving the strength of the stem-cement interlock formed, while preserving the
mechanical integrity (i.e., fracture properties) of the substrate stem material (Jeon et al.,
2012). While clinical studies have assessed the success of various commercially
available elbow prosthesis designs incorporating these surface treatments (Fevang et al.,
2009; Skytta et al., 2009; van der Lugt and Rozing, 2004), there are no known in-vitro
studies that have compared their role in component stability, or provided information on
the failure mechanisms associated with these stem surface treatments.
96
Figure 4.7: Global Stem Motion of Surface Treated Stems in Compression
Global motion of the surface treated stems from 901 cycles (i.e., start of the 10000 N load
step) until failure. The 20 mm and 10 mm beaded stems demonstrated the smallest
magnitudes of stem motion (p < 0.05∞), with the 20 mm plasma spray treatment also
showing less stem motion than the 10 mm plasma spray treatment (p < 0.05¶). Overall,
cobalt chrome stems showed less motion than titanium (p = 0.02).
97
Pistoning of the ulnar component of linked total elbow replacements is one source
of implant loosening, and can occur from axial forces acting at the implant-cement
interface. These axial forces result from dynamic loading of the ulnohumeral joint in
routine daily activities (Amis, 2012; Goldberg et al., 1988; Johnson and King, 2005), or
from impingement of coronoid processes or protruding cement during elbow
hyperflexion, causing distraction forces to occur on the ulnar component (Cheung and
O’Driscoll, 2007). Therefore, in order to compare the effects of different surface
treatments on component stability, axial forces that may contribute to stem pistoning
should be considered. Pistoning forces that occur in-vivo typically act in tension at the
proximal end of the ulnar component; however, the shear forces produced at the stem-
cement interface are the same as those expected if the stems were exposed to similar
compression loads. As such, the purpose of this study was to investigate the role of stem
surface treatment on the stability of titanium and cobalt chrome implant stems, using a
study design incorporating cyclic compression loading to mimic ulnar pistoning at
failure.
Stem surface treatment had a major effect on overall stem stability. When
looking at interface strength and stem motion, the 20 mm beaded treatment showed
greatest survival and least toggle when compared to the other stem surfaces for both
titanium and cobalt chrome stems (Figure 4.6). Beaded treatments have greater surface
deviations with regards to its surface topography, when compared to other stem surfaces.
Based on the diameter of the sintered beads applied to the stem surface, this may allow
for greater infiltration of bone cement onto the stem surface. Failure of the stem-cement
interface would therefore be dependent on the amount of shearing force needed to
overcome each interlock along the stem’s surface. Greater cement infiltration, in terms
of depth and number of interlocking sites, would therefore require greater cyclic loads to
cause stem instability, resulting in interface failure and stem push-out.
For stems that failed catastrophically (i.e., experienced 2 mm push-out),
significant interaction was observed between stem surface treatment and substrate
material for measures of interface strength and toggle. The 20 mm plasma spray cobalt
chrome stem performed better than both of the 10 mm length, surface treatments and
98
smooth cobalt chrome stems (Figure 4.6). For titanium stems, however, the 20 mm
plasma spray treatment performed similar to the 10 mm length, surface treatments and
smooth surfaces. Therefore, although stem surface treatment contributed to improved
interface stability prior to catastrophic failure, this was dependent on the stem substrate
material.
When comparing the interface mechanics for the individual stem-cement
interfaces at catastrophic failure, the smooth stem surfaces appeared to offer no resistance
to the shearing force at the interface, and as such, once the stem-cement bond was
broken, the stem failed rapidly. In comparison, for cobalt chrome stems, the roughened
surface of the 20 mm plasma spray stem provided frictional resistance to the shearing
force at the stem-cement interface, which was greater than that provided by the 10 mm
length, plasma spray and 10 mm length, beaded treatments. Once the shearing force at
the interface became larger than the frictional force, the stems experienced push-out.
Post-testing inspection of the stem surfaces found that the 10 mm beaded stem
was the only surface to experience mechanical damage at failure (Figure 4.3). This
suggests that the number of cycles to failure for the 10 mm beaded treatment may not be
representative of the stem-cement interface strength, but more likely a measure of the
strength of the bond between the beads and stem surface. The strength of this bond is
directly related to the carbon content of the base substrate material, where a lower carbon
content can affect the success of the sintering process, causing a weaker bond between
the stem substrate material and attached beads (Davis, 2003). Although beaded
treatments are likely to contribute to satisfactory interface strengths, as observed from the
20 mm beaded treatments, there may be variability in its performance based on the
success of the sintering treatment during the fabrication process. As such, stringent
standards should be placed on the fabrication process of beaded stem designs, ensuring
adequate material composition of the base substrate metals before application of the
beaded treatments.
For stems that surpassed the staircase region of the loading protocol, and reached
1000 cycles without failure, global stem motion was used to compare the contribution of
99
stem surface treatments to stem motion at the constant 10000 N level. Global stem
motion included motion of the stem with cement, and offered information on the failure
patterns associated with the different stem surface treatments. Overall, beaded stems
experienced less stem motion compared to the plasma spray surfaces. As mentioned
previously, the roughened surface of the plasma spray treatment provided frictional
resistance to the shearing force at the stem-cement interface, and this resistance was
represented by the gradual increase in stem motion prior to catastrophic failure (i.e.,
gradual interface failure). In comparison, catastrophic failure of 10 mm length, beaded
stems was influenced by the bond broken between the beads and the stem surface,
resulting in a stable interface with little stem motion before stem push-out (i.e., rapid
interface failure). When analyzing stem motion for the 20 mm length beaded stems that
did not experience catastrophic failure, it was seen that these values were comparable to
the 10 mm length beaded surfaces, even at the end of the loading protocol (i.e, 10000 N;
25900 cycles). This suggests that the 20 mm length beaded treatment contributed to a
stable stem-cement interface.
Titanium and cobalt chrome are two common stem metals used in ulnar
component designs (Van der Lugt and Rozing, 2004). Each offers its own advantages to
implant systems in terms of biocompatibility, wear resistance and mechanical properties
(Hallab et al., 2004; Navarro et al., 2008). However, application of surface treatments to
the substrate metal can influence the success of the implant system. Clinical studies have
reported that beaded treatments on titanium stems can cause increased rates of component
fracture, believed to be caused by weakening of the metal during the bead sintering
process (Athwal and Morrey, 2006; Jeon et al., 2012). Our study did not evaluate the
strength of the metal stemmed components, but found that the success of the surface
treatments was directly related to stem material type. As mentioned before, variations in
the metal composition, specifically the carbon content of the substrate metal, can directly
affect the strength of the bond formed between the beads and stem during the sintering
process. Therefore, future studies should look into investigating the effect of stem
substrate material composition on the bonding strength of beaded treatments, as well as
the effect of the bead sintering process on the fracture properties of different stem
substrate materials commonly used with elbow prostheses.
100
During elbow flexion, the ulnohumeral joint is under compression/tension, where
resultant forces act upwards onto the distal end of the humerus (Amis, 2012), and varies
depending on the angle of elbow flexion. These joint forces can act on linked total elbow
prostheses in a similar manner, where resultant forces act upwards onto the humeral
component causing pull-out of the connected ulnar component. It is postulated that pull-
out of the ulnar component may also occur from distraction forces produced when the
elbow is hyperflexed past a limit set by an impinging structure (i.e., flange, cement,
bone), creating a fulcrum loading scenario (Jeon et al., 2012), or from carrying a heavy
object during elbow extension, causing forces at the trochlea of up to twenty times the
external load at the hand (Amis et al., 1980). Both axial loading examples create shear
forces along the length of the stem-cement interface, which can cause interface de-
bonding and resultant pistoning of the ulnar component. Our study incorporated a cyclic
compressive load to mimic dynamic shear forces that may cause pistoning of the implant
stem under axial loading, similar to that experienced by the ulnar component. As such,
the loads used for compression testing (i.e., 500–10000 N) were intentionally chosen to
compare the effect of the different surface treatments in a cyclic pistoning scenario,
similar to that caused by resultant joint forces or distraction forces at the ulnohumeral
joint.
From the measures of stem motion as detected from the optical tracking system, it
was observed that the cement directly surrounding the stem contributed to overall stem
motion. This was seen in measures of relative and global stem motion, where the stem
experienced greater motion in the global frame when compared relative to the cement
(Figure 4.5). This may be explained by the creep properties exhibited by bone cement
under dynamic loading (Lewis, 2011). This contribution of cement creep may also
explain the variability seen in our motion results for smooth stems. The stability of
smooth stems was influenced by the stem-cement bond formed at the interface, and as
such, any motion detected prior to failure may have been solely dependent on the visco-
elastic nature of the cement on individual testing days. Bone cement properties could
have also affected the minimal stem motion observed for the 20 mm beaded stems that
completed the testing protocol (i.e., 25900 cycles), since previous work showed that bone
cement becomes stiffer with increasing loading cycles (Verdonschot and Huiskes, 1995).
101
This in-vitro study, to the authors’ knowledge, is the first to compare the effects
of stem surface treatment and substrate material on pistoning of an implant stem under
axial load. The study was successful at showing the contribution of beaded and plasma
spray surface treatments to stem-cement interface stability, as well as able to provide
information on the failure mechanisms associated with these surface treated stems.
Overall, the 20 mm beaded stems offered the greatest stability among all stem surfaces,
and for cobalt chrome stems only, the 20 mm plasma spray stems contributed to
improved stability as well. When comparing mechanisms of catastrophic failure, smooth
stems failed via debonding at the stem-cement interface, beaded stems failed via
debonding of the beaded surface treatment from the stem surface, and plasma spray stems
failed via loss of frictional force between the plasma spray treatment and bone cement. It
is expected that the results from this biomechanical analysis will help to understand the
contribution of surface treatments in component pistoning, and provide information about
the failure mechanisms associated with similar clinical stem designs.
4.5 CONCLUSION
For surface treated stems tested in cyclic compression loading, beaded performed
better than plasma spray treatments on titanium and cobalt chrome stem substrates.
However, plasma spray treatments performed better on cobalt chrome than titanium stem
substrate. Overall, for both beaded and plasma spray treatments stem stability was
improved with a greater surface treatment coverage length.
102
4.6 REFERENCES
Amis, A.A., 2012. Biomechanics of the Elbow. In: Stanley, D., Trail, I. (Eds.), Operative
Elbow Surgery. Churchill Livingstone Elsevier, Edinburgh; New York, pp. 29–44.
Amis, A.A., Dowson, D., Wright, V., 1980. Elbow joint force predictions for some
strenuous isometric actions. Journal of Biomechanics 13, 765–775.
Athwal, G.S., Morrey, B.F., 2006. Revision total elbow arthroplasty for prosthetic
fractures. The Journal of Bone and Joint Surgery; American Volume 88, 2017–2026.
Cheung, E. V, O’Driscoll, S.W., 2007. Total elbow prosthesis loosening caused by ulnar
component pistoning. The Journal of Bone and Joint Surgery; American Volume 89,
1269–1274.
Davis, J.R., 2003. Coatings. In: Handbook of Materials for Medical Devices. ASM
International, Materials Park, OH, pp. 179–194.
Evans, B.G., Daniels, A.U., Serbousek, J.C., Mann, R.J., 1988. A comparison of the
mechanical designs of articulating total elbow prostheses. Clinical Materials 3, 235–
248.
Fevang, B.-T.S., Lie, S.A., Havelin, L.I., Skredderstuen, A., Furnes, O., 2009. Results
after 562 total elbow replacements: A report from the Norwegian Arthroplasty
Register. Journal of Shoulder and Elbow Surgery 18, 449–456.
Goldberg, V.M., Figgie 3rd, H.E., Inglis, A.E., Figgie, M.P., 1988. Total elbow
arthroplasty. The Journal of Bone and Joint Surgery; American Volume 70, 778–
783.
Hallab, N.J., Urban, R.M., Jacobs, J.J., 2004. Corrosion and Biocompatibility of
Orthopedic Implants. In: Yaszemski, M., Trantolo, D., Lewandrowski, K., Hasirci,
V., Altobelli, D., Wise, D. (Eds.), Biomaterials in Orthopedics. Marcel Dekker, Inc.,
New York, pp. 63–91.
103
Jeon, I.H., Morrey, B.F., Sanchez-Sotelo, J., 2012. Ulnar component surface finish
influenced the outcome of primary Coonrad-Morrey total elbow arthroplasty.
Journal of Shoulder and Elbow Surgery 21, 1229–1235.
Johnson, J.A., King, G.J.W., 2005. Anatomy and Biomechanics of the Elbow. In:
Williams, G.R. (Ed.), Shoulder and Elbow Arthroplasty. Lippincott Williams &
Wilkins, Philadelphia, pp. 279–296.
Lewis, G., 2011. Viscoelastic properties of injectable bone cements for orthopaedic
applications: state-of-the-art review. Journal of Biomedical Materials Research, Part
B, Applied Biomaterials 98, 171–191.
Navarro, M., Michiardi, A., Castano, O., Planell, J.A., Castaño, O., 2008. Biomaterials in
orthopaedics. Journal of the Royal Society, Interface 5, 1137–1158.
Skytta, E.T., Eskelinen, A., Paavolainen, P., Ikavalko, M., Remes, V., 2009. Total elbow
arthroplasty in rheumatoid arthritis: a population-based study from the Finnish
Arthroplasty Register. Acta Orthopaedica 80, 472–477.
Van der Lugt, J.C., Rozing, P.M., 2004. Systematic review of primary total elbow
prostheses used for the rheumatoid elbow. Clinical Rheumatology 23, 291–298.
Verdonschot, N., Huiskes, R., 1995. Dynamic creep behavior of acrylic bone cement.
Journal of Biomedical Materials Research 29, 575–581.
104
CHAPTER 5: THE EFFECT OF STEM CIRCUMFERENTIAL
GROOVES ON THE STABILITY AT THE IMPLANT-CEMENT
INTERFACE
Overview: Previous chapters of this thesis discussed the application of beaded and
plasma sprayed treatments for improving stem-cement interface stability in joint
replacement systems; however, results showed variable success of these surfaces. As
opposed to the addition of a treatment or finish to the base stem material, altering stem
design through changing the surface topography by removal of a portion of the base stem
material may offer some advantages. This study compared the effect stem
circumferential grooving, with varying grooved dimensions, on the torsional and axial
stability of cemented stems. Findings from this study show the difference in stability
response between the two grooved designs, and allowed relative comparison of the
circumferential grooves with surface treatments tested in previous chapters of this
thesis.3
5.1 INTRODUCTION
Implant stem design plays an important role in the mechanical stability of implant
systems (Barrack, 2000; Evans et al., 1988). Stem surface modifications, such as the
application of a surface treatment or surface finish, is one such design factor that can
have an influential effect on implant stability (Crowninshield et al., 1998; Jeon et al.,
2012; Scheerlinck and Casteleyn, 2006; van der Lugt and Rozing, 2004). Thus far,
Chapters 3 and 4 of this thesis have explored the effect of stem surface treatment on the
3 A version of this work is in the revision stage of publication: Y.K. Hosein, G.J.W. King, C.E. Dunning
(2013). “The Effect of Circumferential Grooves on the Stability of Cemented Joint Replacement Systems”
Journal of Medical Devices.
105
torsional and axial stability of cemented implant systems. These studies were interested
in the overall effect of stem surface treatments on the mechanical response at the stem-
cement interface, since it was reported that surface treated implants contributed to
reduced rates of loosening in ulnar component systems (Jeon et al., 2012). This clinical
finding seemed reasonable considering the expectation of the altered stem surfaces at the
fixation interface, which would facilitate infiltration of bone cement during the cement
application and curing process. The resulting mechanical connect formed would
therefore provide improved implant anchorage, and resistance to implant loosening.
From the results reported in Chapters 3 and 4, it was confirmed that plasma spray
and sintered bead treatments improved the overall stability of implant systems under
torsional and axial loading; however, their individual contribution to implant stability was
quite variable. Beaded coatings appeared to be more stable than plasma spray coatings,
but depended on the stem substrate material (i.e., cobalt chrome versus titanium). In
addition to this variable stability response, other authors have reported higher fracture
rates associated with beaded implant stems, compared to plasma spray stems, which may
result from the weakening of the stem substrate metal during the bead sintering process
(Athwal and Morrey, 2006; Jeon et al., 2012). As such, the choice between surface
treatment types for improving fixation is not a straightforward decision.
Similarly, there have also been mixed reviews over the use of stem surface
finishes in cemented lower limb implant designs. It is believed that surface finished
stems produce increased cement and/or metal wear at the fixation interface, as a result of
the inevitable micromovement of the stem after implantation (Scheerlinck and Casteleyn,
2006). As such, although the surface finish designs initially provide improved implant
fixation, they may potentially limit the success of the implant system over time.
An alteration in the surface topography of the base stem material, such as the
machining of grooves onto the stem, would eliminate the contribution of an additional
interface (i.e., surface treatment onto the stem) to the stability of the cemented stem
construct, as well as reduce wear caused by surface finishes at the stem-cement interface,
while still allowing for improved mechanical connect and stem fixation during the
106
cementing process. However, the effect on implant stability is unknown. As such, the
purpose of this study was to investigate the effect of circumferential grooves on the
stability of cemented stems under compression and torsional loading. It was
hypothesized that the application of circumferential grooves would not affect torsional
stability, but would have a stabilizing influence in compression, which is of particular
interest to upper limb replacement systems that often fail due to tension-compression
forces at the joint (Cheung and O’Driscoll, 2007).
5.2 MATERIALS AND METHODS
Fifteen metal stems with circular cross-sections (Ø = 8 mm) were custom
machined from cobalt chrome, and made with smooth (n = 5) or circumferentially-
grooved (n = 10) surfaces (Tornier S.A.S., Grenoble, France) (Figure 5.1) (Appendix G).
Individual grooves (i.e., not threaded) were machined along a fixed 20 mm length region
of the stem surface. Groove spacing and depths were either 0.6 mm (n = 5) or 1.1 mm
(n = 5), at a spacing to depth ratio of one.
All stems were potted into square aluminum tubes, to a fixed depth of 20 mm,
using vacuum-mixed polymethylmethacrylate (PMMA) bone cement (Simplex P
,
Stryker
, Kalamazoo, MI, USA). The potted stems were maintained at 22 ºC, and left 24
hours to cure until testing. Subsequently, stems were placed in a materials testing
machine (Instron
8874, Norwood, MA, USA) for application of loads at room
temperature. Compression and torsional tests were done on separate testing days, using
the same implant stems, cleaned with acetone (Caledon Laboratories Ltd., ON, Canada),
and re-potted with new cement.
Compression tests were done using a cyclic staircase loading protocol, similar to
that described in Section 4.2. Cyclic loads fitted a sine wave function, at a frequency of
1.5 Hz. Load targets oscillated between 500 N and an upper limit, which started at
1000 N and increased in increments of 1000 N every 100 cycles to a maximum 10000 N.
107
At 10000 N, cycling continued for a further 25000 cycles, totaling 25900 cycles at the
end of the protocol. Failure was defined as 2 mm of stem push-out, termed catastrophic
failure, or until completion of the loading protocol. For torsional loading of the stems, a
similar protocol to that described in Section 3.2 was incorporated, using a sine wave
pattern to apply cyclic loads at a frequency of 1.5 Hz. Torque values cycled between
0 Nm and an upper limit, which started at 1 Nm and increased in 1Nm increments every
100 cycles to a maximum of 30 Nm. Failure of the implant stem was defined as a rapid
increase in rotation of the stem without resistance, termed catastrophic failure, or until
completion of the protocol.
The custom optical tracking system (Basler Pilot GigE Camera, Ahresnburg,
Germany; Opto Engineering Telecentric Lens, Mantova, Italy; and LabVIEW Vision
Acquisition System, Austin, TX, USA) previously described in Sections 2.2, 3.2 and 4.2
of this thesis, was used to detect motion at the stem-cement interface during compression
and torsional loading. The measure of interface toggle was assessed by the width of the
relative stem motion and rotation graphs (Figure 5.2A and 5.2B). For stems that did not
fail catastrophically before reaching the maximum load/torque (i.e., 10000 N or 30 Nm),
global stem motion was quantified (Figure 5.2C). Global stem motion represented the
displacement of the stem within the camera’s coordinate system (i.e., not relative to the
cement), which occurred during the loading cycles at the maximum load only (indicated
by the starred region in Figure 5.2C).
One-way ANOVAs (α = 0.05) were used to compare the effect of stem surface
condition on interface strength (i.e., load to failure), and stem motion (i.e., both interface
toggle and global stem motion).
5.3 RESULTS
The failures observed were dependent upon the loading mode. In compression,
only the smooth stem experienced catastrophic failure prior to the maximum load, with
108
Figure 5.1: Smooth and Circumferential Grooved Surface Designs Tested in Both
Compression and Torsion
(A) Smooth, (B) 1.1 mm Grooved, and (C) 0.6 mm Grooved. All stems were cemented
to fixed 20 mm depth, as indicated by the region highlighted by the double arrows.
109
both grooved surfaces completing the loading protocol without stem push-out (Figure
5.2A). When subjected to an applied torque, all stem surfaces experienced catastrophic
failure prior to reaching the maximum torque (Figure 5.2B). In both loading modes,
stem motion increased with increasing cycles of the staircase loading protocols, and as
such, stem motion data were normalized to their respective load/torque data for statistical
comparisons.
Overall, the presence of grooves increased the number of compressive loading
cycles achieved prior to failure (p < 0.001) (Figure 5.3A), but had no effect on loading
cycles for torsion (Figure 5.4A).
Under compression, grooved stems experienced less interface toggle prior to
failure compared to smooth surfaces (p < 0.01), with no differences between the two
grooved designs (p = 0.97) (Figure 5.3B). For torsion tests, motion data showed that
grooved 1.1 mm stems experienced the greatest interface toggle prior to catastrophic
failure (p < 0.01), with no differences between the smooth and grooved 0.6 mm surfaces
(p = 0.76) (Figure 5.4B).
When comparing stems that reached the maximum load under compression
without catastrophic failure, grooved 1.1 mm stems showed greater stem motion with
increased cycling, compared to grooved 0.6 mm stems (p = 0.03) (Figure 5.5).
5.4 DISCUSSION
Implant stem designs are constantly evolving in an effort to reduce the incidence
of loosening. Stem surface modification (i.e., application of surface treatment or surface
finish) is one factor of implant design that may reduce the effects of loosening. By
incorporating a roughened stem surface onto cemented implant designs, the bone cement
is allowed to infiltrate the stem surface, providing mechanical fixation between the stem
and cement. For femoral components, surface finished stems are described based on their
roughness value (Ra value) (Verdonschot, 2005), and studies have shown that an
110
Figure 5.2: Stem Motion for Smooth and Circumferential Grooved Stems
Representative graphs of stem motion for smooth, grooved 0.6 mm and grooved 1.1 mm.
(A) Relative stem motion under compression, and (B) relative stem rotation in torsion,
was used to determine interface toggle prior to failure. (C) Global stem motion was used
to determine the change in stem motion with increased number of cycles at the maximum
load, as indicated by the starred region.
111
Figure 5.3: Interface Stability for Smooth and Circumferential Grooved Surfaces in
Compression
(A) Loads required to cause failure and, (B) interface toggle prior to failure. All stems
failed at consistent loads, resulting in zero standard deviation for the measures of load at
failure, as seen in part (A). Smooth stems showed the least stability (p < 0.001**)
compared to the grooved surfaces.
112
Figure 5.4: Interface Stability for Smooth and Circumferential Grooved Surfaces in
Torsion
(A) Torque at failure and, (B) interface toggle prior to failure. No differences were found
in the torque to failure among all stems (p = 0.1), however, grooved 1.1 mm stems
showed greatest interface toggle prior to failure (p < 0.01*).
113
Figure 5.5: Stem Motion with Increased Number of Cycles
Stem motion during the loading cycles at maximum compression load, for grooved
0.6 mm and grooved 1.1 mm stems. Grooved 1.1 mm stems showed increased stem
motion compared to the grooved 0.6 mm (p = 0.3).
114
increased roughness value contributed to increased interface strength; however, there
were implications of interface wear and abrasion once debonding had occurred
(Crowninshield et al., 1998; Scheerlinck and Casteleyn, 2006). Similarly for ulnar
components, stem surface treatments such as sintered beads and plasma spray
incorporated onto cemented stem designs have shown improved clinical success (Jeon et
al., 2012; van der Lugt and Rozing, 2004); however, in-vitro testing of these surfaces in
previous chapters of this thesis has demonstrated variable outcomes dependent on stem
substrate material. As such, this chapter examines a different design concept for the stem
surface, with the aim of improving implant fixation without the contradictory effects
contributed by roughened stem surfaces. Therefore, the purpose of this study was to
investigate the effect of circumferential grooves on the stability of implant stems under
compression and torsional loads.
The hypotheses for this study were accepted. Under compression, both grooved
surfaces significantly improved the interfacial strength compared to the smooth surface.
When comparing interface strength for torsional loading, no statistical differences were
found among the grooved and smooth surfaces, as expected, given the direction of the
grooves relative to the applied torque. Despite this, there was a trend towards increased
interface strength with the application of grooves (Figure 5.4A). Thus, with regards to
interface strength, it appears that the circumferential design of the grooves created
mechanical interlocks along the length of the stem-cement interface and increased the
stem surface contact with cement, offering superior resistance to shear forces at the
interface. This interfacial resistance would be dictated by the mechanical properties, and
in particular, the shear strength of the integrated bone cement. In comparison, smooth
stems created an adhesive-type bond with the cement at the interface, and provided some
frictional resistance to interface forces. Once these bonding and frictional forces were
overcome by the interface shear forces, the smooth stems experienced failure.
Toggle results for compression loading showed that both grooved surfaces
experienced little interface toggle compared to the smooth surface, suggesting superior
interface stability and resistance in compression. When comparing the grooved stems at
the constant 10000 N load with increasing load cycles, the grooved 1.1 mm stem
115
experienced increased stem motion prior to failure. Similarly, results of interface toggle
under torsional loading showed that the grooved 1.1 mm surfaces experienced the largest
magnitudes of toggle compared to the other stem surfaces. This may be explained by the
size and number of grooves associated with the grooved 1.1 mm stem, which would
accommodate a larger volume of cement integration onto the stem surface. As such, the
mechanical properties of bone cement would have a greater influence on the interface
motion of the grooved 1.1 mm stem surface compared to grooved 0.6 mm, and smooth
surfaces. Although bone cement is used as a connector, it is a viscoelastic material
(Lewis, 2011), and can cause increased motion of the stem without leading to interface
failure.
When comparing the results of this study to those previously reported for surface
treated implants in Sections 3.3 and 4.3 of this thesis, circumferential grooves performed
comparable to, or better than, surface treated stems in compression. Although they did
not improve torsional stability, other stem design factors such as stem cross-sectional
shape (Kedgley et al., 2007) and curvature (Berzins et al., 1993; Evans et al., 1988), may
work in combination with the circumferential grooves to provide increased torsional
stability when incorporated into implant stem designs. In addition, applying longitudinal
grooves in combination with the circumferential design may work in synergy to improve
both axial and torsional stability; however, further studies are needed to confirm this
hypothesis.
Cobalt chrome was the stem material chosen for testing stem circumferential
grooves within this study, since results from Sections 3.3 and 4.3 of this thesis showed
that this stem substrate material contributed to the variable mechanical response of the
beaded treatments in compression and torsional loading. Therefore, the present study
aimed to determine the stability response of stem circumferential grooves on cobalt
chrome stems, to consider its possible use as an alternative surface design to the beaded
treatment on cobalt chrome stems. In addition, it was not expected that stem material
would influence the mechanical response of the respective circumferential grooved
surfaces; therefore, only one stem material was utilized in the study design.
116
Cementing technique plays an important role in the fixation mechanics of joint
replacements (Faber et al., 1997; Iesaka et al., 2003). Likewise, the properties of the
bone cement during preparation (i.e., viscosity, temperature) can affect the mechanical
connect formed at the stem-cement interface. This study used low viscosity bone cement
for fixation of the implant stems, allowing for better infiltration of cement into the
grooved surfaces and improving the mechanical interlock formed once the cement was
cured. For the grooved 0.6 mm stems, in particular, low viscosity bone cement would
likely ensure sufficient infiltration into the smaller groove spacing. As such, it is
important to consider cement consistency during implantation, to allow for proper
interdigitation of cement with the grooved stem surface.
Cement shrinkage is another property of bone cement that can affect the success
of cemented interfaces (Bishop et al., 1996; Eveleigh, 2001; Kwong and Power, 2006;
Lennon and Prendergast, 2001; Lewis, 1997; Orr et al., 2003). In particular, the cement
used in this study (i.e., Simplex P) has been shown to have approximately 6 percent
decrease in its initial volume during curing (Kwong and Power, 2006). Shrinkage of the
cement is initiated by polymerization and temperature effects of the cemented construct
(Draenert and Draenert, 2005). However, reports on the effect of cement shrinkage at the
stem-cement interface have been variable, ranging from the formation of gaps (Bishop et
al., 1996), to shrinkage and compression of the cement onto the stem (Draenert and
Draenert, 2005). While this study did not investigate the effects of cement shrinkage on
the interface formed with the grooved surfaces, it is important to consider, since the
overall stability of the grooved stems were shown to be influenced by the properties of
the integrated bone cement. Therefore, future studies should look into methods for
quantifying cement shrinkage at the stem-cement interface.
Although the grooved surface improved implant stability, the effect of the
grooved edges within the cement should be acknowledged. Sharp edges at the stem-
cement interface can act as stress risers (Crowninshield et al., 1980; Evans et al., 1988),
initiating cement fracture. Incorporating a rounded edge on the grooved corners may
prevent stress fractures in the cement. While this study did not investigate the effects of
117
the grooved stem surface on fractures within the bone cement, this should be addressed in
future finite element studies.
Another limitation of adding grooves to the stem surface is their potential effect
on stem strength. The presence of the grooves reduces the minimum cross sectional area
of the stem. In addition, the notches created by the grooves on the stem surface may act
as stress risers, reducing the overall strength of the stem. However, by controlling the
groove dimensions and adding fillets (i.e., rounded internal corners) these effects may be
mitigated.
Stem circumferential grooves improved implant stability under compression, with
no significant effects in torsion. Overall, grooved 0.6 mm stems experienced less stem
motion compared to the grooved 1.1 mm stems. Therefore, introducing circumferential
grooves onto implant stem surface designs may reduce the effects of implant loosening
caused by compressive forces.
5.5 CONCLUSION
When comparing grooved stems, the grooved 0.6 mm stems showed improved
stem stability compared to grooved 1.1mm stems in compression, with a trend towards
improved stability in torsion as well. This alternative surface topography may address
the concerns associated with surface treated and surface finished stems, and potentially
reduce the cost of the stem fabrication process.
118
5.6 REFERENCES
Athwal, G.S., Morrey, B.F., 2006. Revision total elbow arthroplasty for prosthetic
fractures. The Journal of Bone and Joint Surgery; American Volume 88, 2017–2026.
Barrack, R.L., 2000. Early failure of modern cemented stems. The Journal of
Arthroplasty 15, 1036–1050.
Berzins, A., Sumner, D.R., Andriacchi, T.P., Galante, J.O., 1993. Stem curvature and
load angle influence the initial relative bone-implant motion of cementless femoral
stems. Journal of Orthopaedic Research 11, 758–769.
Bishop, N.E., Ferguson, S., Tepic, S., 1996. Porosity reduction in bone cement at the
cement-stem interface. The Journal of Bone and Joint surgery; British volume 78,
349–356.
Cheung, E. V, O’Driscoll, S.W., 2007. Total elbow prosthesis loosening caused by ulnar
component pistoning. The Journal of Bone and Joint Surgery; American Volume 89,
1269–1274.
Crowninshield, R.D., Brand, R.A., Johnston, R.C., Milroy, J.C., 1980. An analysis of
femoral component stem design in total hip arthroplasty. The Journal of Bone and
Joint Surgery; American Volume 62, 68–78.
Crowninshield, R.D., Jennings, J.D., Laurent, M.L., Maloney, W.J., 1998. Cemented
femoral component surface finish mechanics. Clinical Orthopaedics and Related
Research 355, 90–102.
Draenert, K., Draenert, Y., 2005. The Three Interfaces. In: Breusch, S.J., Malchau, H.
(Eds.), The Well-Cemented Total Hip Arthroplasty. Springer Berlin Heidelberg,
New York, pp. 93–102.
119
Evans, B.G., Daniels, A.U., Serbousek, J.C., Mann, R.J., 1988. A comparison of the
mechanical designs of articulating total elbow prostheses. Clinical Materials 3, 235–
248.
Eveleigh, R., 2001. Mixing systems and the effects of vacuum mixing on bone cement.
British Journal of Perioperative Nursing 11, 132–140.
Faber, K.J., Cordy, M.E., Milne, A.D., Chess, D.G., King, G.J., Johnson, J.A., 1997.
Advanced cement technique improves fixation in elbow arthroplasty. Clinical
Orthopaedics and Related Research 334, 150–156.
Iesaka, K., Jaffe, W.L., Kummer, F.J., 2003. Effects of preheating of hip prostheses on
the stem-cement interface. The Journal of Bone and Joint Surgery; American
Volume 85, 421–427.
Jeon, I.H., Morrey, B.F., Sanchez-Sotelo, J., 2012. Ulnar component surface finish
influenced the outcome of primary Coonrad-Morrey total elbow arthroplasty.
Journal of Shoulder and Elbow Surgery 21, 1229–1235.
Kedgley, A.E., Takaki, S.E., Lang, P., Dunning, C.E., 2007. The effect of cross-sectional
stem shape on the torsional stability of cemented implant components. Journal of
Biomechanical Engineering 129, 310–314.
Kwong, F.N.K., Power, R. a, 2006. A comparison of the shrinkage of commercial bone
cements when mixed under vacuum. The Journal of Bone and Joint Surgery; British
Volume 88, 120–122.
Lennon, A.B., Prendergast, P.J., 2001. Evaluation of cement stresses in finite element
analyses of cemented orthopaedic implants. Journal of Biomechanical Engineering
123, 623–628.
Lewis, G., 1997. Properties of acrylic bone cement: state of the art review. Journal of
Biomedical Materials Research 38, 155–182.
120
Lewis, G., 2011. Viscoelastic properties of injectable bone cements for orthopaedic
applications: state-of-the-art review. Journal of Biomedical Materials Research, Part
B, Applied Biomaterials 98, 171–191.
Orr, J.F., Dunne, N.J., Quinn, J.C., 2003. Shrinkage stresses in bone cement. Biomaterials
24, 2933–2940.
Scheerlinck, T., Casteleyn, P.P., 2006. The design features of cemented femoral hip
implants. The Journal of Bone and Joint Surgery; British Volume 88, 1409–1418.
Van der Lugt, J.C., Rozing, P.M., 2004. Systematic review of primary total elbow
prostheses used for the rheumatoid elbow. Clinical Rheumatology 23, 291–298.
Verdonschot, N., 2005. Philosophies of stem designs in cemented total hip replacement.
Orthopedics 28, s833–s840.
121
CHAPTER 6: THE EFFECT OF STEM SURFACE TREATMENT
AND MATERIAL ON THE STABILITY OF JOINT REPLACEMENT
SYSTEMS SUBJECTED TO BENDING LOADS
Overview: Joint loading and resultant interface stresses play a key role in the stability of
cemented implants. Previous chapters of this thesis have focused on the effect of
torsional and axial loads on implant stability, and discussed the role of stem surface
treatments in resisting the resultant shear stresses at the stem-cement interface. Bending
loads at the joint, however, result in both normal and shear stresses at the stem-cement
interface, which can have a varied effect on the interface’s mechanics. As such, this
chapter investigates the effect of stem surface treatment on the interface stability of
cemented implants subjected to bending loads.
6.1 INTRODUCTION
The influence of stem surface treatments on the stability of cemented implant
systems is dependent on their ability to resist interface stresses caused by joint loading.
However, depending on the loading mode type, resultant stresses at the stem-cement
interface can vary. These variable interface stresses can lead to different mechanisms of
stem debonding, and resultant implant loosening.
Chapters 3 and 4 of this thesis have already explored the role of surface treated
implant stems on the axial and torsional stability of implant systems. In particular, these
studies showed similar results with regards to the performance of sintered beads and
plasma spray treatments, on titanium and cobalt chrome stems, in both loading conditions
(Sections 3.3 and 4.3). The results seem reasonable considering the resultant stresses that
occur at the stem-cement interface due to these loading modes. Shear stresses act
122
concentric to stem-cement interface in torsional loading, and parallel to the interface
under axial loading.
In addition to torsional and axial loads occurring at the replaced joint, bending
loads are also apparent, as a result of joint forces acting perpendicular to the longitudinal
axis of the implant (Bergmann and Graichen, 2010; Bergmann et al., 2007, 2004, 1993;
Guerra, 2004; Kutzner et al., 2010; Westerhoff et al., 2009). These forces result in
normal interface stresses that act perpendicular to the stem-cement interface, in addition
to shear interface stresses that act parallel to the stem surface.
Normal interface stresses that occur during bending of the stem can be tensile or
compressive in nature, depending on the location of the interface relative to the applied
perpendicular load. These interface stresses can lead to gradual stem debonding, which is
more likely to occur at the region of the stem-cement interface that experiences tensile
stresses (Huiskes and Schouten, 1980; Huiskes, 1985). Debonding at the interface can
subsequently lead to instability of the stem within the cement. As such, implant systems
subjected to bending loads should also be considered when investigating the overall
effect of surface designs on implant stability.
Therefore, the purpose of this study was to investigate the effect of stem surface
treatment on the stability of titanium and cobalt chrome implant stems under bending
loads. It was hypothesized that smooth stems would offer inferior stem fixation
compared to the surface treated stems, forming a weak bond at the stem-cement interface
that would be susceptible to debonding and stem instability under bending loads. In
comparison, it was expected that the surface treated stems would show minimal motion at
the stem-cement interface, due to the mechanical resistance to interface tensile and shear
stresses. In addition, it was expected that titanium stems may experience greater stem
motion as a result of its lower modulus of elasticity than cobalt chrome.
123
6.2 MATERIAL AND METHODS
Fifty implant stems (Ø = 8 mm) were custom machined from titanium (n = 25)
and cobalt chrome (n = 25) by Tornier S.A.S. (Montbonnot Saint Martin,France). For
each stem substrate material, five stems retained their smooth stem surface, and twenty
stems had standard commercially employed stem surface treatments applied by Orchid
Bio-Coat (Southfield, MI, USA); 20 mm length beaded treatment (n = 5), 10 mm length
beaded treatment (n = 5), 20 mm length plasma spray treatment (n = 5), and 10 mm
length plasma spray treatment (n = 5), as previously described in Chapter 4.
The fixation method used in this study was similar to that in previous chapters,
where all stems were centralized, and potted to a fixed cement depth of 20 mm using
vacuum-mixed PMMA bone cement (Simplex P
, Stryker
, Kalamazoo, MI, USA). The
stems were cured for 24 hours in a temperature controlled environment, and subsequently
tested in a materials testing machine (Instron
8874, Norwood, MA, USA). Custom
fixturing was used to secure the stem at the base of the materials testing machine. The
cemented stem in the aluminum tube was positioned on its side, so that the longitudinal
axis of the stem was parallel to the base of the materials testing machine. A ball bearing
attached to the end of the load applicator was used to apply compressive forces to the
centre of the stem head, creating a bending moment about the stem head (Figure 6.1). A
cyclic compressive staircase loading protocol was used for testing of the stems; they were
pre-loaded to 10 N in compression, and cycled between 10 N and upper load limit value,
which started at 50 N and increased in 50 N increments every 100 cycles, to a maximum
of 1000 N in compression (limited to ensure there was no permanent deformation to the
stem external to the cement).
Stem-cement interface motion was observed using the optical tracking system
(Basler Pilot GigE Camera, Ahresnburg, Germany; Opto Engineering Telecentric Lens,
Mantova, Italy; and LabVIEW Vision Acquisition System, National Instruments, Austin,
TX, USA) described in previous chapters of this thesis (Section 2.2. 3.2, 4.2, and 5.2).
The displacement between markers placed on the stem and cement (Figure 6.1) was used
124
Figure 6.1: Experimental Set-up used for Bending Tests of Implant Stems
Cyclic compressive loads were applied to the stem head, inducing bending of the stem
within the cement construct (side view shown in schematic at top right). The camera was
used to track markers placed on the stem and cement (shown in inset image), and the
distance between the markers was defined as relative stem motion.
125
to determine the relative stem motion during loading. From the graphs of relative stem
motion over time, interface stability was quantified by “maximum interface toggle”,
which represented the maximum width of the relative stem motion graphs during loading.
An additional measure of stem stability was the “offset stem motion”, which was a
measure of the average deviation of the stem from its original position at the start of
loading (Figure 6.2).
Visual analysis of the cemented stem construct was done post-testing, to
determine whether the stem was loose within the cement mantle, or if the cement mantle
appeared to be compromised, during mechanical testing.
Two-way analyses of variance with post-hoc Student-Newman-Keuls tests
(α = 0.05) were used to examine the role of stem surface treatment and stem material on
implant stability under bending loads, using measures of interface toggle and offset stem
motion.
6.3 RESULTS
Post-testing visual inspection of the cemented stem construct found that all stems
remained wedged within the cement mantle, without indications of stem loosening. In
addition, there were no signs of damage or fracture to the exposed regions of bone
cement. However, it was observed that stems appeared displaced within the cement,
where analysis of the stem from the proximal and distal ends of the construct showed the
stem tilted from its central axis, which was originally in a horizontal configuration
(Figure 6.3). This tilt was quantified by the measure of offset stem motion (Figure 6.2).
Statistical analysis of maximum interface toggle found an overall effect of stem
surface treatments (p = 0.033), where the smooth stems (0.074 ± 0.023 mm) experienced
greater interface toggle compared to the 20 mm beaded treatment (0.047 ± 0.012 mm)
(p = 0.018). However, there were no differences found among the other surface
126
Figure 6.2: Stem Motion of Surface Treated Stems under Bending Loads
Representative graphs of relative stem motion for surface treated stems with titanium and
cobalt chrome substrate materials. Maximum interface toggle was defined as the
maximum width of the relative motion graphs ( ), and offset stem motion was defined as
the average deviation of stem motion from the starting position, shown by dashed line
and (¥).
127
Figure 6.3: Proximal and Distal Views of Stems Post- Bending Tests
Post-testing inspection of the (A) proximal, and (B) distal views of the cemented implant
stems found that the stems deviated from their central axes (yellow dotted line), and
appeared offset within their cement mantles.
128
Figure 6.4: Maximum Interface Toggle for Surface Treated Stems in Bending
The 20 mm length beaded treatments experienced less toggle than the smooth stems
(p = 0.018) (as indicated by *), with no differences found between stem materials
(p = 0.587).
129
treatments (p > 0.05), as well as no difference in interface toggle between stem materials
(p = 0.587) (Figure 6.4).
For offset stem motion, differences among the surface treatments were found
(p = 0.041), where smooth stems (0.051 ± 0.023 mm) experienced greater stem deviation
within the cement mantle than the 20 mm beaded treatment (0.031 ± 0.008 mm). In
addition, there was no effect of stem material on offset stem motion (p = 0.577) (Figure
6.5).
6.4 DISCUSSION
Joint loads acting perpendicular to the implant stem can induce bending of the
cemented stem construct. Throughout various joints of the body, these loads have been
shown to range between 0.05 to 5 times body weight during routine daily activities
(Bergmann and Graichen, 2010; Bergmann et al., 2007, 2004, 1993; Guerra, 2004;
Kutzner et al., 2010; Westerhoff et al., 2009), resulting in varied bending moments acting
about the stem head. During bending of the cemented stem, loads are transferred across
the stem-cement interface creating regions of tensile, compressive and shear stresses
along the interface. These stresses can initiate debonding of the stem from the cement,
leading to implant micromotion and subsequent implant loosening.
Chapters 3 and 4 of this thesis discussed the role stem surface treatments played
in interface stability under torsional and axial loads, and in particular their contribution to
resisting shear stresses at the stem-cement interface. With bending loads, however, the
response of surface treatments to resist normal stresses (i.e., tensile and compressive) at
the interface may vary. As such, the purpose of this study was to investigate the effect of
stem surface treatment on the stability of titanium and cobalt chrome stems under
bending loads.
130
Figure 6.5: Offset Motion for Surface Treated Stems in Bending
The 20 mm length beaded stems demonstrated less offset stem motion than the smooth
stems (as indicated by *), with no differences found between stem materials (p = 0.577).
131
Analysis of interface stability found that stem surface treatment had an overall
effect on interface stability under bending loads, where smooth stems showed greater
interface toggle compared to the 20 mm beaded treatment (Figure 6.4). As discussed
previously within this thesis, smooth stems create a chemical bond with the cement at the
interface, which is more susceptible to interface stresses compared to the mechanical
interlock created by the 20 mm beaded treatment. As such, it is likely the smooth stems
may have experienced debonding resulting in greater interface instability compared to the
20 mm beaded treatment.
The results of interface toggle observed among the surface treatments (i.e, 20 mm
beaded, 20 mm plasma spray, 10 mm beaded, and 10 mm plasma spray treatment) could
be explained by the mechanism of stability expected for the surface treatments under
bending loads. Surface treated stems promote interdigitation of bone cement onto the
stem surface, which reduces the chances of stem debonding due to normal and shear
stresses acting at the stem-cement interface during stem bending. However, the improved
interface fixation offered by the surface treated stems may result in the contribution of
bone cement to interface toggle measurements. As such, similar interface toggle
observed among the surface treated stems may likely be influenced by the mechanical
properties of the interlocked bone cement. Despite the non-difference in the results, the
overall trend of interface toggle for each of the surface treatments may be indicative of
the contribution of the individual surface treatments to interface stability.
This contribution of the mechanical properties of bone cement was also observed
in the measure of offset stem motion, where results found that all stems deviated from
their original position within the cement mantle during loading, and smooth stems
appeared to show greater deviations compared to the 20 mm beaded treatment (Figure
6.5). Depending on the stem-cement interface condition (i.e., fixed or debonded), the
load transfer across the interface can vary (Huiskes and Schouten, 1980; Huiskes, 1985).
For smooth stems that experience stem debonding, tensile interface stresses are no longer
able to be transmitted, and as such, compressive stresses at stem-cement interface are
likely to increase. This increase in compressive interface stresses can cause increased
loading of the surrounding bone cement. The bone cement consequently exhibits its own
132
mechanical response to loading, which has been shown to be viscoelastic in nature
(Lewis, 2011, 1997; Saha and Pal, 1984), and results in creep of the cement and
subsequent drift of the stem within the cement mantle.
Although post-testing inspection of the cemented stem construct found that the
stems were offset from their original position, all stems appeared to be securely wedged
within the cement mantle. This is important to note since surgical analysis of the
cemented stem constructs may assume that a secure stem is indicative of a stable stem
construct; however, our results have shown that although stems were not physically loose
within the cement mantle, all stems appeared to experience some interface toggle and
stem deviation under bending. While these values may appear small in magnitude, they
are likely to increase with continuous exposure to interface stresses, as expected in the
case of in-vivo joint loading.
The loading protocol used in this study was chosen to represent typical bending
moments expected at the shoulder, elbow, hip and knee joints (Bergmann et al., 2007,
1993; Guerra, 2004; Kutzner et al., 2010), taking into consideration the dimensions of the
cemented stem construct. As such, a cyclic staircase protocol was used with bending
loads between 50–1000 N, which created bending moments of 1–20 Nm about the stem
head. This protocol also allowed direct comparison of the various surface treatments
within the time-frame required for 24 hours testing of the cemented stem samples.
The optical system used for measuring interface stability proved useful for
determining stem-cement motion based on analysis of the exposed regions of the stem
and cement. However, considering the influential role that bone cement plays in stem
stability under bending stresses, analysis of the full length of the cemented stem construct
may be needed to fully understand the interface mechanics involved in implant bending.
As such, future studies should look into experimental or computational methods that
allow analysis of the entire cemented stem construct.
The findings from this study demonstrated that smooth stems experienced greater
instability compared to the 20 mm beaded treatments under bending loads, with no
differences found among the other surface treatments. Overall, stem substrate material
133
had no effect on stem stability. Additionally, results from this study suggest that bone
cement plays a significant role in the mechanical response of cemented implant stems
exposed to bending loads; however, future studies will be needed to confirm this.
6.5 CONCLUSION
Stem surface treatments did not improve implant stability under cyclic bending
loads; however, overall stem motion appeared to be influenced by creep or motion of the
surrounding bone cement.
134
6.6 REFERENCES
Bergmann, G., Graichen, F., 2010. Realistic loads for testing hip implants. Bio-medical
Materials and Engineering 20, 65–75.
Bergmann, G., Graichen, F., Bender, A., Kääb, M., Rohlmann, A., Westerhoff, P., 2007.
In vivo glenohumeral contact forces--measurements in the first patient 7 months
postoperatively. Journal of Biomechanics 40, 2139–2149.
Bergmann, G., Graichen, F., Rohlmann, A., 1993. Hip joint loading during walking and
running, measured in two patients. Journal of Biomechanics 26, 969–990.
Bergmann, G., Graichen, F., Rohlmann, A., 2004. Hip joint contact forces during
stumbling. Langenbecks Archives of Surgery 389, 53–59.
Guerra, S.M., 2004. Design and development of a total elbow prosthesis to quantify
ulnohumeral load transfer. (M.E.Sc Thesis) University of Western Ontario, pp 85-
108.
Huiskes, H., Schouten, R., 1980. Effect of interface loosening on the stress distribution in
intramedullary fixated artificial joints. In: American Society of Mechanical
Engineers (Ed.), Advances in Bioengineering. New York, pp. 213–216.
Huiskes, R., 1985. Properties of the stem-cement interface and artificial hip-joint failure.
In: The Bone-Implant Interface. p. 86.
135
Kutzner, I., Heinlein, B., Graichen, F., Bender, A., Rohlmann, A., Halder, A., Beier, A.,
Bergmann, G., 2010. Loading of the knee joint during activities of daily living
measured in vivo in five subjects. Journal of Biomechanics 43, 2164–2173.
Lewis, G., 1997. Properties of acrylic bone cement: state of the art review. Journal of
Biomedical Materials Research 38, 155–182.
Lewis, G., 2011. Viscoelastic properties of injectable bone cements for orthopaedic
applications: state-of-the-art review. Journal of Biomedical Materials Research, Part
B, Applied Biomaterials 98, 171–191.
Saha, S., Pal, S., 1984. Mechanical properties of bone cement: a review. Journal of
Biomedical Materials Research 18, 435–462.
Westerhoff, P., Graichen, F., Bender, A., Halder, A., Beier, A., Rohlmann, A.,
Bergmann, G., 2009. In vivo measurement of shoulder joint loads during activities
of daily living. Journal of Biomechanics 42, 1840–1849.
136
CHAPTER 7: THE USE OF MICRO-COMPUTED TOMOGRAPHY
(µ-CT) IMAGING TO VISUALIZE AND QUANTIFY MOTION AT THE
STEM-CEMENT INTERFACE WITH AN APPLIED BENDING
MOMENT
7.1 INTRODUCTION
Stem-cement interface motion is a useful measure for investigating the stability
response of cemented implant systems exposed to joint loading. It provides information
regarding the localised motion of the stem relative to the adjacent cement, but can be
influenced by the mechanical properties (i.e., creep) of the connecting/interface cement.
Bone cement is viscoelastic in nature, and as such, exhibits displacement under loads
(Lewis, 2011, 1997; Saha and Pal, 1984). The extent of this response depends on several
factors, including the type of loading that occurs at the joint (i.e., compression/tension,
torsion, bending) (Lewis, 2011, 1997; Saha and Pal, 1984; Verdonschot and Huiskes,
1995), and can affect the overall stability at the stem-cement interface. Therefore,
investigation into the effects of loading on the response at the internal stem-cement
interface can offer insight into the interface mechanics that dictate implant stability.
Bending loads, in particular, offer a unique loading response at the stem cement
interface, which has been discussed in previous sections of this thesis (i.e., Section 6.1
and 6.4). Depending on the direction of the load, the interface experiences regions of
tensile and compressive stresses (Huiskes and Schouten, 1980; Huiskes, 1985).
However, the effect of these bending stresses on the displacement response at the
interface is not intuitive. Based on the results from Section 6.3, it was found that all
implant stems experienced stem motion under bending loads; however, post-testing
inspection found that the stems remained securely held within their cement mantles. As
such, it was hypothesized that bone cement creep contributed to the observed stem
motion, rather than motion resulting from interface loosening. This inference could not
137
be validated, however, since the optical tracking system was only capable of measuring
motion of the exposed regions of the stem and cement (i.e., where the stem exited the
cement), with no direct indication of the motion occurring along the length of the stem
within the cement mantle. Therefore, an appropriate tool was needed to access the
internal stem-cement interface during bending, to determine motion along the entire
length of the cemented stem.
Micro-computed tomography (µ-CT) imaging is a useful visualization tool that
has been used in orthopaedic research to investigate the internal properties of bone and
implanted systems (Bernhardt et al., 2006; Blok et al., 2013; Stock, 2009a; Teeter, 2012;
Waarsing et al., 2005). The high resolution of the acquired images (i.e.,50–100 µm)
(Stock, 2009b), allows detection of microstructure details within the object. This has
proven useful for studies, including the investigation of bone morphology (Bernhardt et
al., 2006; Blok et al., 2013; Waarsing et al., 2005), and wear characteristics of joint
replacement components (Teeter, 2012). It was hypothesized that the detailed images
provided by the µ-CT technology may also be suitable for detecting micromotion at the
interfaces of an implanted system.
As such, this study aimed to use µ-CT imaging to observe changes at the internal
stem-cement interface during loading. Taking into account the observations from
Chapter 6 of this thesis, where ambiguity remained regarding the influence of bending
loads on the mechanical response at the internal stem-cement interface, this chapter also
intended to address those uncertainties. Therefore, the purpose of this study was to
investigate the influence of bending loads on internal stem-cement interface motion,
using micro-computed tomography (µ-CT) imaging to visualize the interface.
138
7.2 MATERIALS AND METHODS
7.2.1 TESTING SPECIMENS
Nine implant stems were tested, each machined from titanium by Tornier (France)
to consist of a circular cross-section body (Ø = 8 mm), with full length smooth surface.
All stems were potted to a fixed depth of 10 mm in aluminum tubes using an orthopaedic
bone cement (Simplex P, Stryker, Kalamazoo, MI) mixture incorporating miniature steel
beads (Metaltec Steel Abrasive Co., Canton, MI) (Ø = 500microns) (please see Appendix
K for details on mixing method). The beads were used as cement markers in subsequent
CT images of the stem-cement construct. The appropriate bead to cement ratio was
determined to ensure the original flexural modulus of the bone cement was best
maintained, while still allowing a reasonable dispersion of beads within the cement
mantles (see Appendix K for details concerning the selection of this ratio).
Once the stems were potted, they were left to cure at a constant 22 ºC
temperature. All stems were allowed 24hrs of curing time before subsequently being
placed into a custom built loading device for application of bending loads.
7.2.2 CUSTOM-BUILT LOADING DEVICE
The custom built loading device, previously designed for application of bending
loads to ulna bone specimens, was modified to incorporate the cemented stem constructs
(Figure 7.1) (see Appendix M for details of the loading device). Using the initial design
concept and the major components of the device, minor modifications were made to
accommodate the application of bending loads to the cemented stem constructs. The
design specifications ensured that the device was made of CT-compatible materials, and
able to fit within the bore of the CT scanner. In addition, the final prototype was
evaluated to ensure its durability during the application of bending loads.
139
The mechanical components of loading device incorporated a stepper motor
(Robbins and Myers, Electrocraft, Willis, TX) connected to an offset bearing cam, which
was further attached to a pivoted carbon fibre beam (Figures M.1 and M.2). During one
revolution of the motor, the cam connection actuated the pivoted carbon fibre beam in a
see-saw motion, with approximately 1 mm of vertical displacement. An aluminum high
accuracy S-beam load cell (LCR Series, Omega, Stamford, CT) was mounted at the top
of the carbon fibre beam, above which a ball bearing was placed for the application of
load to the head of the cemented stem construct. An aluminum jig secured onto the
device’s platform was used to hold the cemented stem construct in a horizontal
orientation above the carbon fibre beam (Figure 7.1). This configuration allowed forces
to be applied perpendicular to the cemented stem construct, resulting in a bending
moment applied to the implant stem.
7.2.3 LOADING AND IMAGING PROTOCOL
Stems were loaded in a combination of static and dynamic conditions, from a
minimum load of 50 N to a maximum load of 960 N, in the following sequences: 1)
monotonic increase, 2) cyclically loaded at a rate of 0.9 Hz for 1000 cycles (to initiate a
fatigue type response within cemented stem construct), and 3) monotonic increase
(Figure 7.2).
For each of the static load levels (i.e., pre- and post-dynamic testing), CT scans of
the cemented stem were obtained at the unloaded and loaded condition (Figure 7.2),
using a micro-CT scanner with voxel spacing of 154 µm (eXplore Ultra, GE Healthcare,
London, Canada). The scans were acquired at an x-ray potential of 120 kVp, with 40 mA
tube current, over the duration of 8 seconds. The resultant image volume was 1024 mm x
1024 mm x 360 mm.
140
Figure 7.1: Custom-Built Loading Device used for Application of Bending Loads
(A) The loading device was designed to be secured onto the CT bed, and fit within the
bore of the scanner. For application of bending loads, (B) the cemented stem was
secured in a horizontal orientation into an aluminum holding jig, which was fixed to the
base of the loading platform. The platform housed a stepper motor connected to an offset
cam, which was attached to a pivoted carbon fibre beam. During one revolution of the
motor, the beam translated vertically in a sinusoidal displacement pattern. A load cell
attached to the top of the beam recorded the load applied to the stem head during vertical
translation of carbon fibre beam. The applied load induced bending within the cemented
stem construct, as shown in the inset schematic. The region highlighted by the dashed
red box indicates the volume of the bending device which was included in scanner’s field
of view.
141
Figure 7.2: Schematic of Loading and Imaging Protocol used for Static and
Dynamic Tests
All stems were loaded in a combination of static and dynamic conditions from a
minimum load of 50 N to a maximum load of 960 N, in the following sequences: 1)
monotonic increase, 2) cyclically loaded at a rate of 0.9 Hz for 1000 cycles, and 3)
monotonic increase. CT scans were acquired for the unloaded and loaded condition of
the pre- and post- dynamic sequences, as highlighted by the red dashed boxes, and used
for comparison of stem-cement motion during loading. A cyclic sequence was
incorporated between the two monotonic sequences to initiate a fatigue type loading
response within the cement mantle.
142
7.2.4 DATA ANALYSIS
Scans of the cemented stems were reconstructed, and converted from .vff’s (i.e.,
image files) to .stl’s (i.e., surface geometry files) using MicroView 3D Image Viewer and
Analysis Tool (Parallax Innovations, Ilderton, ON). The surface files were subsequently
imported into Geomagic Qualify (Geomagic, Morrisville, NC) for analysis of stem-
cement interface motion (see Appendix N for details of analysis). For each stem, the
unloaded and loaded surfaces of the pre dynamic scans were compared to determine the
response of the stem and the surrounding beads embedded within the cement to the
applied static loads. The same analysis was done using the post-dynamic scans. In
addition, the unloaded surfaces of the pre- and post-dynamic scans were compared to one
another, to determine offsets in stem position as a result of dynamic loading.
This comparison involved registering the surfaces from the two scans, using the
square aluminum tube containing the stem as the fixed point of reference, and
determining the deviations between the two surface profiles (Figure 7.3A). 2D sections
along the length of the implant (Figure 7.3B) were used to analyse changes in the stem
and bead surfaces. Measurements of stem surface deviation were obtained from the
central 2D profile, while measurements of bead motion were obtained from 2D profiles
located within a 2 mm range of this central 2D profile (such that the nearest bead was
identified). The vertical change in the stem surface between the unloaded and loaded
conditions was quantified (Figure 7.3 C), and the vertical and horizontal changes in bead
surfaces were analyzed. However, due to inconsistent deformations in a few of the bead
surfaces (see Appendix O), a stringent criteria was applied to analyze bead surface
deviation. That was, bead deviation was only recorded if contralateral edges on the bead
surface experienced similar quantities of deviations.
Measurements of stem motion were taken at four specific locations along the
stem: at the point of load application (i.e., stem head), the beginning of the stem shaft
(i.e., uncemented shaft), the beginning of the cement mantle (i.e., cemented shaft #1), and
the end of the cement mantle (i.e., cemented shaft #2) (Figure 7.4). The beads embedded
within the cement were analyzed based on their location relative to two specific regions
143
Figure 7.3: Surface Deviation Analysis with Geomagic®
Motion of the stem and cemented beads was measured from deviation of their respective
surfaces between unloaded and loaded scans. Shown above is the loaded stem with beads
embedded within the cement. (A) The 3D color plot shows the change in stem position
from the unloaded to loaded scan, with red (+ve) and blue (-ve) regions showing greatest
stem surface deviation. (B) A 2D section along the length (highlighted by the plane) of
the cemented implant was used to analyze changes in the stem and bead surfaces. (C)
From the sectioned profiles, the vertical change in the stem surface between the unloaded
(red profile) and loaded (black profile) conditions was measured both before and after the
cyclic loading phase. The load was applied to the head of the stem using a ball bearing,
as seen in Figure 7.1. (Note: The aluminum tube containing the cemented stem was
removed from the image for better visualization of the stem-cement interface).
144
within the cement (Tables 7.1–7.3). Region 1 of the cement represented the length
between cemented shaft #1 and the midpoint along the length of the cement mantle.
Region 2 represented the cement length between cemented shaft #2 and the same
midpoint of the cement mantle.
A one-way repeated measures ANOVA, with Student-Newman-Keuls post-hoc
tests (α = 0.05), was used to compare the magnitudes of stem motion from the pre- and
post-dynamic scans, thereby determining the effect of cyclic bending loads on the
internal mechanical response of the cemented stem constructs. Bead deviations,
however, were inconsistent among stems, and as such, provided only a qualitative
measure of cement motion (i.e., no statistical analyses were performed).
7.3 RESULTS
Inspection of the stems post-testing (i.e., after completion of the three loading
sequences) found that all stems remained securely wedged within their cement mantles,
with no signs of stem loosening. Statistical analysis comparing the change in stem
motion between the pre- and post- dynamic CT images, found no difference in the motion
of the stem head (p = 0.373), or either of the uncemented (p = 0.162) and cemented
regions of the stem shaft (p > 0.05) (Figure 7.4). However, the average motion observed
for each of stem regions showed a trend towards reduced stem motion following the
application of dynamic loads (Figure 7.4).
When observing motion along the entire length of the stem during loading, it was
found that stems experienced tilting about a pivot point within the cement mantle. This
pivot point varied in position among the stems, shifting from the center to the proximal
and distal regions (i.e., regions 1 and 2, respectively) of the cement (Tables 7.1 and 7.2).
Analysis of bead motion within the cement found that some beads experienced
vertical and horizontal translation, while others remained fixed (Tables 7.1 and 7.2).
145
Figure 7.4: Average Stem Motion during Loading for the Pre- and Post-dynamic
Scans
Comparison of stem motion (i.e., head, uncemented shaft and cemented shaft) between
the unloaded and loaded scans, for the pre- and post-dynamic sequences, found that there
was no difference in stem motion (p > 0.05) as a result of dynamic loading. However,
there was a trend towards decreased stem motion after exposure to cyclic loads, as
observed from the reduced displacement in the post-dynamic sequence.
146
Table 7.1: Pre-dynamic Cemented Stem and Bead Motion
Displacement along the length of the cemented mantle for the cemented stem shaft and
beads embedded within the cement prior to the cyclic loading sequence. The arrows (↑, ↓)
indicate the direction of stem and bead motion. The presence of beads was not consistent
between regions, or among the stem samples tested, therefore, dashes (-) represent the
absence of a bead within the region, while zero (0) represents an identifiable bead that did
not displace.
Stem
#
Cemented
Shaft
(mm)
Cemented
Shaft
(mm)
# 1 # 2 Region 1 Region 2 Region 1 Region 2
1 0.040 ↑ 0.059 ↓ - 0.039 ↓ 0 0.045 ←
2 0.021 ↑ 0.010 ↓ 0.036 → - 0.050 ↑ -
3 0.034 ↑ 0.025 ↓ - 0.041 → - 0
4 0.067↑ 0.063 ↓ 0.031 ↑ 0 - 0.134 ←; 0.078 ↓
5 0.033 ↑ 0.026 ↓ 0 - 0 -
6 0.055 ↑ 0.012 ↓ 0 0.025 ↓ 0 -
7 0.050 ↑ 0.044 ↓ - 0.032 →, 0.062 ↓ 0 0.036 ←; 0.047 ↓
8 0.042 ↑ 0.035 ↓ - - 0 0
9 0.040 ↑ 0.039 ↓ - - 0 0
Bead Motion (Above Stem) (mm) Bead Motion (Below Stem) (mm)
147
Table 7.2: Post-dynamic Stem and Bead Motion
Displacement along the length of the cemented mantle for the cemented stem shaft and
beads embedded within the cement following the cyclic loading sequence. The arrows (↑,
↓) indicate the direction of stem and bead motion. The presence of beads was not
consistent between regions, or among the stem samples tested, therefore, dashes (-)
represent the absence of a bead within the region, while zero (0) represents an identifiable
bead that did not displace.
Stem
#
Cemented
Shaft
(mm)
Cemented
Shaft
(mm)
# 1 # 2 Region 1 Region 2 Region 1 Region 2
1 0.029 ↑ 0.050 ↓ - 0.040 ↓; 0.037→ 0 0
2 0.059 ↑ 0.002 ↓ 0 - 0 -
3 0.049 ↑ 0.045 ↓ - 0.020→ - 0.049 ←
4 0.037 ↑ 0.005 ↓ 0 0 - 0.031 ↓
5 0.052 ↑ 0.026 ↓ 0 - 0 -
6 0.058 ↑ 0.011 ↓ 0 0 0 -
7 0.038 ↑ 0.012 ↓ - 0 0.027 ↑ 0
8 0.039 ↑ 0.027 ↓ - - 0 0
9 0.004 ↑ 0.042 ↓ - - 0 0.028 ↓
Bead Motion (Above Stem) (mm) Bead Motion (Below Stem) (mm)
148
Table 7.3: Offset Stem and Bead Motion
Displacement along the length of the cemented mantle for the cemented stem shaft and
beads embedded within the cement obtained when comparing the unloaded scans from
pre- and post-dynamic testing. The arrows (↑, ↓) indicate the direction of stem and bead
motion. The presence of beads was not consistent between regions, or among the stem
samples tested, therefore, dashes (-) represent the absence of a bead within the region,
while zero (0) represents an identifiable bead that did not displace.
Stem
#
Cemented
Shaft
(mm)
Cemented
Shaft
(mm)
# 1 # 2 Region 1 Region 2 Region 1 Region 2
1 0.014 ↑ 0.018 ↓ - 0 0 0.071←
2 0.016 ↑ 0.003 ↓ 0 - 0 -
3 0.009 ↑ 0.014 ↓ - 0 - 0
4 0.040 ↑ 0.042 ↓ 0 0 - 0.119←; 0.027↓
5 0.010 ↑ 0.010 ↓ 0 - 0 -
6 0.015 ↑ 0.006 ↓ 0 0 0 -
7 0.029 ↑ 0.012 ↓ - 0 0.080 ← 0.068←
8 0.009 ↑ 0.025 ↓ - - 0.035 → 0
9 0.007 ↑ 0.013 ↓ - - 0 0
Bead Motion (Above Stem) (mm) Bead Motion (Below Stem) (mm)
149
When comparing offset stem and bead motion between the pre- and post-dynamic
unloaded scans, it was observed that all cemented constructs experienced offset motion as
a result of dynamic testing, with some constructs showing greater motion than others
(Table 7.3).
7.4 DISCUSSION
The clinical success of implant systems is dependent, in part, on the stability
provided by the connection formed at the stem-cement interface (Barrack, 2000; Mohler
et al., 1995). However, the ability of this connect to resist forces that occur at the
interface, may be influenced by the mechanical properties of the individual interface
materials (i.e., bone cement and metal). Bone cement, in particular, is known to exhibit
a time-dependent displacement response when loaded (Lewis, 2011, 1997; Saha and Pal,
1984). Therefore, in order to fully understand the interface mechanics that occur during
joint loading, the contribution of individual interface components to overall stem stability
needs to be explored.
Chapter 6 of this thesis discussed the unique interface stress response initiated by
bending loads. From those results, it was hypothesized that bone cement displaced under
interface stresses, and thereby affected the overall stability of the stem. This contribution
is important to consider since displacement of cement can cause motion of the stem
without loosening, eventually leading to implant misalignment. As such, a thorough
investigation into the effect of bending loads on the response of the internal stem-cement
interface was needed.
Visual inspection of the cemented stem constructs post-testing found that all
stems were secured within their cement mantles, with no signs of stem instability or
loosening. This would suggest that any motion observed for the loaded stems may have
been contributed to by the displacement of the surrounding bone cement. Statistical
analysis found no difference in stem motion between the pre- and post-dynamic loaded
scans, implying that exposure to short-term dynamic loading did not cause a significant
150
increase in stem instability. However, from the graphs of stem motion between the pre-
and post- dynamic conditions, there was a trend towards reduced stem motion with
application of dynamic loads (Figure 7.4). This may be explained by the mechanical
response of bone cement to cyclic loads. If stem motion was dictated by displacement of
the surrounding cement, it would seem reasonable that the stems experienced less motion
as a result of short–term stiffening of bone cement with increased number of loading
cycles (Verdonschot and Huiskes, 1995).
When comparing the motion of the stem along the cemented interface, a distinct
pattern was observed in the motion path of the stem. During loading the stems appeared
to experience tilting about a pivot point within the cement mantle, suggesting that
displacement of the surrounding cement facilitated stem motion (Tables 7.1 and 7.2). For
some stems this pivot point occurred at the midpoint along the length of the mantle,
where a positive displacement of the stem in the proximal region of the cement
reciprocated to negative displacement in the distal region. For other stems this pivot
point shifted to the proximal or distal region of the cement mantle. The difference in the
location of the pivot point may offer an explanation into the fixation condition along the
length of the interface during bending. In a fully-fixed stem condition the normal
stresses across the interface would be higher at the proximal and distal ends of the
cemented construct, with a middle region of minimal stress transfer (Huiskes and
Schouten, 1980; Huiskes, 1985). This would result in increased stem motion at the
proximal and distal ends of the cemented stem. In a partially-fixed stem condition,
however, the stress transfer across the interface would change (Huiskes and Schouten,
1980; Huiskes, 1985), shifting the pivot point to the fixed region of the interface. It is
important to note though, that a well-fixed or partially-fixed interface condition in
bending may not be indicative of immediate stem loosening, since all stems remained
securely wedged within the cement mantle. Over time, however, the partially-fixed
condition may result in increased stem motion and complete debonding at the interface.
These interface conditions, particularly for the pre-dynamic scans, demonstrate the
variability of bond formation created at the stem-cement interface during cement fixation
of smooth stem surfaces.
151
The specific response of cement to bending loads was observed in the motion of
beads within the cement. While all beads did not experience displacements, the beads
that showed motion under loading did so in accordance with the direction of the applied
load. This observation further verified that the cement experienced some displacement
under bending loads. It is believed that the observed bead movement within the cement
may have been dictated by the bead position along the length of the stem, along with the
fixation condition at the interface. Therefore, for a fully fixed stem condition that
experiences pivoting about a central position within the cement mantle, the beads at the
proximal and distal regions of the cement (i.e., Regions 1 and 2) may likely show
displacement as a result of normal stress transfer across the interface. In comparison, for
a partially fixed condition where the pivot point shifts within the mantle, bead motion
may only occur in the region of the fixed interface. However, due to the inconsistency of
bead distribution within the cement, among the various stems, this hypothesis could not
be fully validated. As such, future work should look into embedment methods that allow
for controlled positioning of markers within the cement mantle.
The beads used in this study were made of steel, with an approximate density of
8 g/cm3, which made them easy to detect within the cement (approximate density of bone
cement = 1.2 g/cm3
(Saha and Pal, 1984)). However, this difference in density may also
result in the beads not being able to move freely with the bone cement. It may be likely
that motion of the bulk cement could have occurred around the embedded beads, rather
than initiate bead motion. This would explain the variation in bead displacements
observed with the cemented constructs. A less dense bead material may have shown
greater motion with the cement during loading. In addition, because of the high density
of the steel bead, it is possible that metal artifacts could have been introduced into the
CT-images. These artifacts would have subsequently affected the surface profiles used
for bead motion analysis (Appendix O). Therefore, it is advised that future studies using
this technique should incorporate a suitable cement marker made of a less dense material
than the surrounding bone cement.
Analysis of the stem and bead motion during loading utilized a method that
measured deviations in stem and bead surfaces between the unloaded and loaded scans.
152
While this method was useful and allowed relative comparison of stem and bead motion,
it depended on the quality of the surfaces created during the conversion of image to
surface files. Image parameters used for conversion were standardized for all surfaces;
however, variability in image quality between scans could have affected the resultant
surface quality.
Smooth surface, titanium stems were used for testing within this study. Smooth
stems were chosen based on the results from Chapter 6, which showed an overall trend
towards increased stem motion compared to the other stems, and significant increase in
stem motion compared to the 20 mm beaded treatment. Additionally, smooth surfaces
showed large variability in their stability results, and as such, it was believed that
investigation of the internal stem-cement interface of smooth stems under bending
moments could offer insight into the interface mechanics associated with these stem
surfaces. Titanium was chosen as the material for CT-testing because it is a low density
metal, and was not expected to significantly contribute to metal artifacts within the
image. However, based on pilot testing comparing the surface deviation between two
consecutive scans of unloaded stems (Appendix O), it was found that some artifacts may
have still occurred, showing small deviations in the stem surface without the application
of a load. Despite this, titanium stems were still useful for CT-imaging tests, compared
to other high density implant metals, and offered satisfactory image quality and results
during testing.
This study was able to interpret the effect of cyclic loads on the mechanical
response of the interface by comparing the motion of the stem and beads during static
loading, both before and after the application of dynamic loads (Figure 7.4), where load
magnitudes (i.e., 50–980 N) and rate (i.e., 0.9 Hz) were chosen in close accordance with
the bending protocol used in Chapter 6.4 While this method was useful for the purposes
of this study, it does not give a true representation of the interface’s response during
4 The loading rate of 0.9 Hz was selected for use in future studies, aimed at acquiring scans during dynamic
testing. This rate would be required to ensure appropriate gating of the loading device with the acquisition
of images from the CT scanner (Armitage et al., 2012).
153
dynamic loading. A better representation would involve the comparison of scans
obtained during cyclic loading. This could be done using a unique imaging technique
described by Armitage et al., where a gated CT-image acquisition method would be used
to acquire scans of the moving stem (Armitage et al., 2012). The technique would
involve image acquisitions of the cyclically loaded stems, and retrospective sorting of the
projection data, to reconstruct images at appropriate phases of the cyclic protocol. This
would result in a series of reconstructed images of the cemented stem construct at
different time points of the cyclic waveform, which could be used to determine the
motion of the stem and beads within the cement during one complete cycle of the
dynamic protocol. The loading fixture, dynamic protocol and analysis methods
developed in this study were designed such that they could be readily employed for such
a test. Future studies will look at incorporating this technique to measure dynamic
motion of the stem and beads within the cemented stem construct.
This is the first known in-vitro study to incorporate µ-CT imaging to determine
real-time interface motion during loading, by incorporating an active loading device
within a µ-CT scanner, to obtain images of loaded cemented stem constructs. The study
showed that stem stability was not significantly altered by exposure to dynamic bending
loads, but found that cement displacement contributed to stem motion during bending.
7.5 CONCLUSION
From µ-CT image analysis along the length of the internal stem-cement interface,
all stems experienced tilting motion about a fixed pivot point within the cement mantle
during the application of bending loads, and the bone cement adjacent to the stem
demonstrated motion as well.
154
7.6 REFERENCES
Armitage, S.E.J., Pollmann, S., Detombe, S., Drangova, M., 2012. Least-error projection
sorting to optimize retrospectively gated cardiac micro-CT of free-breathing mice.
Medical Physics 39, 1452–1461.
Barrack, R.L., 2000. Early failure of modern cemented stems. The Journal of
Arthroplasty 15, 1036–1050.
Bernhardt, R., Scharnweber, D., Müller, B., Beckmann, F., Goebbels, J., Jansen, J.,
Schliephake, H., Worch, H., 2006. 3D analysis of bone formation around titanium
implants using micro computed tomography (µCT). In: Bonse, U. (Ed.), Proceedings
SPIE 6318, Developments in X-ray Tomography, 631807. pp. 1–10.
Blok, Y., Gravesteijn, F. a, van Ruijven, L.J., Koolstra, J.H., 2013. Micro-architecture
and mineralization of the human alveolar bone obtained with microCT. Archives of
Oral Biology 58, 621–627.
Huiskes, H., Schouten, R., 1980. Effect of interface loosening on the stress distribution in
intramedullary fixated artificial joints. In: American Society of Mechanical
Engineers (Ed.), Advances in Bioengineering. New York, pp. 213–216.
Huiskes, R., 1985. Properties of the stem-cement interface and artificial hip-joint failure.
In: The Bone-Implant Interface. p. 86.
Lewis, G., 1997. Properties of acrylic bone cement: state of the art review. Journal of
Biomedical Materials Research 38, 155–182.
Lewis, G., 2011. Viscoelastic properties of injectable bone cements for orthopaedic
applications: state-of-the-art review. Journal of Biomedical Materials Research Part
B, Applied Biomaterials 98, 171–191.
Mohler, C.G., Callaghan, J.J., Collis, D.K., Johnston, R.C., 1995. Early loosening of the
femoral component at the cement-prosthesis interface after total hip replacement.
The Journal of Bone and Joint Surgery; American Volume 77, 1315–1322.
155
Saha, S., Pal, S., 1984. Mechanical properties of bone cement: a review. Journal of
Biomedical Materials Research 18, 435–462.
Stock, S.R., 2009a. Cellular or Trabecular Solids. In: Micro Computed Tomography-
Methodology and Applications. Taylor & Francis Group, Boca Raton, FL, pp. 171–
213.
Stock, S.R., 2009b. Introduction. In: Micro Computed Tomography- Methodology and
Applications. Taylor & Francis Group, Boca Raton, FL, pp. 1–8.
Teeter, M., 2012. Assessment of Wear in Total Knee Arthroplasty Using Advanced
Radiographic Techniques. (Ph.D Thesis) The University of Western Ontario, pp 1-
36
Verdonschot, N., Huiskes, R., 1995. Dynamic creep behavior of acrylic bone cement.
Journal of Biomedical Materials Research 29, 575–581.
Waarsing, J.H., Day, J.S., Weinans, H., 2005. Longitudinal micro-CT scans to evaluate
bone architecture. Journal of Musculoskeletal & Neuronal Interactions 5, 310–312.
156
CHAPTER 8: GENERAL DISCUSSION AND CONCLUSIONS
Overview: This chapter re-examines the original objectives and hypotheses proposed in
Chapter 1 of this thesis to assess whether each of them were successfully accomplished
and proven. Overall strengths and limitations of this body of work are discussed, along
with potential future directions for similar implant biomechanics studies. Finally, the
overall significance of this type of basic science research in the field of orthopaedics is
highlighted.
8.1 SUMMARY
Implant loosening remains the most common mode of joint replacement failure
(Australian Orthopaedic Association, 2010; New Zealand Orthopaedic Association,
2010), with mechanical loosening resulting from exposure of implant systems to joint
loads. Basic implant biomechanics studies, such as those presented within this thesis, can
offer an explanation to clinical questions surrounding mechanical loosening. This thesis
focused specifically on the implant’s stem surface design and its role in the clinical
success of implant systems. Within the literature, surface modified, cemented implants
have been reported to experience reduced rates of loosening (Jeon et al., 2012); however,
the mechanical response associated with these stem surface designs have not been
explored. Therefore, the overarching purpose of the series of studies presented within
this thesis was to investigate the effect of clinically-relevant stem surface designs on the
mechanical stability at the stem-cement interface, under various loading modes, using in-
vitro biomechanical analysis tools to assess stem micromotion.
Chapter 1 provided the basic concepts and background knowledge that was used
to prepare the study designs for subsequent chapters within the thesis. This chapter
introduced joint replacement systems, and factors that may contribute to their clinical
success. In addition, various tools used in the biomechanical analysis of implant motion
157
were detailed, with focus on the use of optical systems as a non-invasive method for
interface analysis. This chapter concluded with the expected objectives and hypotheses
for each study within this thesis.
Initial focus was the development of an appropriate measurement tool to assess
interface micromotion (Objective #1). Chapter 2 described the development and
validation of a custom optical tracking system that was capable of measuring
displacements on the order of micrometers. The hardware components of the optical
system were carefully chosen to meet these requirements. The software used for
displacement measurements incorporated a color thresholding method that detected the
(x,y) coordinates of coloured markers placed on specific landmarks, and calculated the
relative distances between them. The system was validated by comparing motion applied
by a micrometer screw gauge, with that measured by the optical tracking system. It was
determined that the optical system showed agreement with the micrometer screw gauge
for measurements between 0.005 mm and 0.250 mm, based on the scatter of the Bland-
Altman plots, and these measurements were reliable based on the results of intraclass
correlation coefficients (ICC’s) greater than 0.99. Therefore, it was concluded that the
optical system was satisfactory for its application in measuring interface motion within
subsequent chapters of this thesis (Hypothesis #1 accepted).
Using the optical tracking system described in Chapter 2, the stability response of
clinically-relevant stem surface treatments was investigated under torsional loads
(Objective #2, Chapter 3). Cemented titanium and cobalt chrome implant stems
consisting of smooth, plasma spray and beaded surface treatments were mechanically
tested using a cyclic staircase torsional loading protocol, and motion between the stem
and cement was observed. Overall, it was found that stem surface treatments improved
the torsional stability of stems compared to smooth stem surfaces (Hypothesis #2
accepted), but the individual stability responses of plasma spray and beaded treatments
were dependent on the stem substrate material. For cobalt chrome stems, the 20 mm
length plasma spray treatment showed greater interface strength than the 20 mm length
beaded treatments, with no differences in stem rotation prior to failure. However, for
titanium stems, no difference in interface strength was found between the 20 mm length
158
plasma spray and 20 mm length beaded treatment, but the plasma spray stem experienced
more stem rotation prior to failure. In addition, titanium stems showed greater torques at
failure, but greater stem rotation prior to failure, compared to cobalt chrome stems.
The axial stability of surface treated stems was also considered. This was of
particular interest to the upper limb literature, since it has been reported that ulnar
components of linked elbow replacement systems commonly experienced failure via stem
‘pistoning’, or pull-out (Cheung and O’Driscoll, 2007). For cemented implants,
improvement in the mechanical connect at the stem-cement interface, by incorporation of
stem surface treatments, may result in improved resistance to ulnar component pistoning.
Therefore, Chapter 4 of this thesis investigated the effect of stem surface treatment and
material on the stability of implant stems under compression (Objective # 3). Similar to
Chapter 3, it was found that stem surface treatment improved the axial stability of
implant stems compared to smooth surfaces, and this effect was dependent on stem
material (Hypothesis #3 accepted). It was found that the 20 mm length beaded treatments
showed greatest interface strength, with the 20 mm length plasma spray treatment on
cobalt chrome stems also showing improved interface strength, compared to the 10 mm
beaded and plasma spray treatments. In addition, the 20 mm length beaded and plasma
spray treatments showed least interface motion, with the 10 mm length beaded and
plasma spray treatments on titanium stems showing reduced interface motion as well.
When comparing stem material, titanium and cobalt chrome stems performed similar to
one another under compression.
Considering the variable response of stem surface treatments in Chapters 3 and 4
of this thesis, where results found that the success of surface treatments were dependent
on the stem substrate material to which they were applied, Chapter 5 aimed to investigate
an alternative stem surface design (i.e., circumferential grooves) and its contribution to
the axial and torsional stability of cobalt chrome implant stems (Objective #4). It was
determined that the application of 0.6 mm and 1.1 mm circumferential grooves to the
stem surface (i.e., separation and depth) offered improved axial stability compared to
smooth stems, but offered similar stability response in torsion (Hypothesis #4 accepted).
When comparing groove dimensions, no differences were found in interface strength and
159
interface motion between the 0.6 mm and 1.1 mm circumferential grooved surfaces under
compression, however, grooved 1.1 mm stems showed greater stem motion than the
grooved 0.6 mm stems in torsion. In addition, it was suggested that motion observed for
the grooved stems was influenced by the creep properties of the infiltrated bone cement,
since the 1.1 mm grooved stems showed greater overall stem motion than the 0.6 mm
grooved stems.
Besides torsional and axial loads, joint replacement systems are also exposed to
bending loads, which result from joint forces acting perpendicular to the length of the
implant stem. Bending loads produce both shear and normal stresses along the length of
the stem-cement interface, with regions of interface experiencing both tensile and
compressive stresses, dependent on the direction of the applied load. This unique loading
profile can result in varied mechanical response from surface treated implant stems.
Chapter 6 of this thesis investigated the role of stem material and surface treatment on the
stability of implant stems exposed to bending loads (Objective #5). Based on
observations of implant stem motion, it was found that titanium and cobalt stems showed
similar stability responses under bending loads, with only the 20 mm length beaded
treatments experiencing less interface motion than the smooth stem (Hypothesis #5
rejected). In addition, stem motion was thought to be influenced by the displacement, or
creep, of the surrounding bone cement, since all stems demonstrated offset stem motion
within their cement mantle during loading, without any indication of stem loosening.
This was further supported by the observed tilt of the stem away from its central axis, in
post-testing analysis.
Taking into consideration the contribution of bone cement creep to the stability
response of the stems tested under bending in Chapter 6, it was determined that analysis
of the full length of the stem-cement interface was needed to determine the response of
both the stem and cement under bending loads. Chapter 7 of this thesis used µ-CT
imaging to access the internal stem-cement interface, and facilitated motion analysis
along the full length of the interface (Objective #6). A custom built loading device was
used to apply a static load to a smooth implant stem, while allowing real-time scanning of
the loaded stem. Beads embedded within the cement were used as cement markers to
160
detect motion of the cement during loading. Scans of the loaded implant stem, obtained
before and after exposure to cyclic dynamic testing, detected stem and cement embedded
bead motion during loading (Hypothesis # 6 accepted). Motion of the stem appeared to
occur about a pivot point within the cement, where stem tilting similar to that observed in
Chapter 6 was noted. However, the location of this pivot point within the cement mantle
was variable among the stems tested, suggesting different fixation conditions at the
interface during loading. Although motion of the cement embedded beads were not
consistent among stems, this preliminary result was able to prove that cement motion
does occur with implant stems during loading.
8.2 STRENGTHS AND LIMITATIONS
Specific strengths and limitations for each of the individual studies within this
thesis have already been discussed within their respective chapters, but the consolidation
of these studies showed overall strengths and limitations as well.
The series of basic science studies presented within this thesis allowed
investigation of implant stability, dependent of the effects of stem surface design and
material. The studies tested circular cross-section implant stems (Ø = 8 mm), with
altered surface topography along a 20 mm or 10 mm region of the implant stems. While
these stems were not representative of a specific implant design, they allowed relative
comparison of stem surface treatment and material. In particular, circular cross-section
implant stems were chosen since they were shown to exhibit the least resistance to
torsional loads (Kedgley et al., 2007). In addition, since this study was only interested in
motion occurring at the stem-cement interface, stems were potted in standard sized
aluminum tubes, without incorporation of variable bone specimens. The use of
aluminum tubes controlled the cement volume and interfaces, but contributed to thicker
cement mantles, approximately twice the size of those used clinically. The increased
thickness could have contributed to the bone cement creep observed in the studies;
however, larger mantle sizes allowed testing of implant stems without compromise to the
161
surrounding bone cement. Thinner cement mantles may have resulted in failure of the
bone cement, subsequently affecting the stability results of the surface treated implants.
Therefore, although this bench top study design may have limited the clinical scope of
these studies, it facilitated exclusive testing of the stem design features without
contribution from other variables. Furthermore, given that the cement mantle thickness
was consistent across all stems tested, the relative differences would still be reasonable
given the repeated measures design of the studies.
The in-vitro environment used for testing implant stems within this thesis was
different from that expected at the joints in the body. In particular, the temperature and
physical environment did not represent typical in-vivo conditions. All stems were potted
and tested at approximately 22 ºC; however, temperatures within the human body are
higher. Temperature differences can affect the rate of polymerization of bone cement,
with higher temperatures increasing the polymerization rate, and may result in a different
loading response of bone cement than that observed in this thesis. In addition, the
anatomical joint is exposed to joint fluids and blood that are likely to affect the adhesion
properties of stem with the cement. Therefore, while the stems tested within this thesis
were compared relative to one another in the same testing environment, it is important to
note that in-vivo conditions could have resulted in different implant stability outcomes.
The overall aim of this thesis was to investigate the effect of stem surface design
under various loads, to determine the stability response of these designs for improved
resistance to loosening. Therefore, studies within this thesis tested stems under
individual loading modes using custom designed fixtures to secure and accurately align
the cemented stems within the materials testing machine, minimizing the effects of off-
axis loading, as monitored by the six degree of freedom load cell of the materials testing
machine. Although loading that occurs at the anatomical joint acts in a combined state,
the individual contribution of each of these components are important for a thorough
understanding the mechanical response of surface treatments under specific loads. This
is especially useful for studies like the one presented in Chapter 4, where its study design
was developed based on the specific failure response (i.e., pistoning) exhibited by ulnar
components at the elbow joint (Cheung and O’Driscoll, 2007). As such, testing the effect
162
of surface treatments to resist these dominant axial loads was imperative for providing
knowledge regarding failure mechanisms associated with these surface designs at the
ulnohumeral joint.
The magnitudes of loads used for testing of implant stems in torsion, compression
and bending, were chosen based on a review of the literature reporting joint loads
measured by instrumented implants at the hip, knee and shoulder. However, the
maximum loads chosen for the individual testing protocols were on the higher end of
loads expected at the joints. While this did not simulate typical joint loads in daily
activities, it allowed relative comparison of various surface treatments, by initiating a
failure response in the majority of stem surfaces. In addition, the loading protocol used
for all implants was cyclic, staircase in nature. This protocol was specifically chosen to
ensure that all stems were subjected to a range of dynamic loads, which mimicked
repetitive loading experienced by the human joint, while still allowing comparison of
interface strength based of loads required to cause failure.
In- vitro implant biomechanics have used stem motion as an indicator of implant
stability. Few studies however, have compared stem motion relative to the surrounding
bone cement. This measurement, however, is more representative of interface motion,
and allows analysis of implant motion independent of motion occurring within the
surrounding fixturing. The optical tracking system (Chapter 2) and CT measurement
technique (Chapter 7) used in this thesis facilitated micromotion measurements at the
level of the stem-cement interface. This was advantageous to studying localized interface
motion, and further provided additional information regarding the contribution of the
bone cement to implant stability.
8.3 FUTURE DIRECTIONS
The work done within this thesis has provided the framework for investigation of
interface mechanics associated with various stem surface designs. Considering the
influential role that stem surface designs play in the mechanical connect at the stem-
163
cement interface, future work should analyze the response of these varying connections
on the loading response at the interface. This could be done using finite element analysis,
by creating models of the stem-cement interface (Mann et al., 1991), with interface
conditions governed by the different mechanical connects associated with the various
surface designs. These analyses would offer information about the resultant stresses and
strains induced at the interface as a result the varied stem surfaces connects, and more
specifically, the response of the interfacial cement to these varied surface designs.
In addition to utilizing modeling techniques to assess the loading response at the
stem-cement interface, embedment of strain gauges within the cement could be also used
to experimentally analyze the effect of stem surface designs on interface strains (Fetterly,
2012). This would allow stability testing of surface modified implant stems, with
simultaneous analysis of the strain response at the interface. The combination of these
measurements may prove useful in future implant studies to provide a comprehensive
analysis of the interface mechanics that govern the success of cemented stem surface
designs.
The alternative stem surface design (i.e., stem circumferential grooves) tested in
Chapter 5 of this thesis, showed promising results with regards to its stability response
under torsional and axial loads. This concept of machining patterns onto the stems, rather
than incorporating an additional material interface, can limit the variable mechanical
response associated with stem surface treatments and coatings, while improving the
interface connect and overall stem stability. However, this notion of removing material
from the stem needs to be further tested, since alteration to the stem design can also affect
the mechanical strength of the stem. Therefore, future studies should look into
alternative stem surface designs that involve machining patterns onto the stem surface,
and compare these stems for improved interface stability, and mechanical response under
various loading conditions.
Considering the influential role that bone cement played in the stability response
of implant stems throughout this thesis, future studies should also look into the
mechanical behavior of bone cement specific to implant stability. Chapter 7 of this thesis
164
introduced a preliminary method using CT-imaging to observe motion within the cement
mantle during loading of implant stems; however, this technique is still in its rudimentary
stage and needs to be further developed and explored. In addition to acquiring scans
during static loads, the methodology could also be used to analyze stem and cement
motion under dynamic loading, by incorporating unique CT-acquisition techniques to
obtain scans at specific time points of the dynamic loading cycle. The resultant series of
images could then be analyzed to determine interface motion for one full cycle of the
dynamic waveform. The ground work for analyzing stem-cement interface motion from
CT images has already been developed within Chapter 7, therefore future studies could
look at implementing this methodology in the analysis of dynamic interface motion.
Additionally, an improved CT-compatible loading device could be designed to facilitate
testing of implant stems under various loading modes (i.e., compression, torsion,
bending, and combined loading), for analysis of the internal stem-cement interface
response to these individual loading conditions.
8.4 SIGNIFICANCE
In conclusion, the studies presented within this thesis are the first known in-vitro
studies to assess the effect of clinically-relevant stem surface treatments on the stem-
cement interface stability, under various loading modes. The results support the theory
that stem surface designs affect the mechanical connection at the stem-cement interface,
with surface treatments generally providing increased resistance to interface loosening.
Overall, the findings from this thesis have provided valuable information to the literature
regarding the interface mechanics and failure mechanisms associated with stem surface
treatments and machined grooved surfaces, as well as demonstrated the influential role of
stem material on the success of surface treated implants. Across all loading modes,
beaded treatments applied to titanium stems, and plasma spray treatments applied to
cobalt chrome stems, improved interface stability and strength when large surface
treatment areas were employed. In addition, the machining of circumferential grooves
onto the stem surface improved interface strength in compression, with no influence in
torsion. However, smaller groove dimensions resulted in improved stem stability. It is
165
expected that this knowledge can be applied to future cemented implant designs, to
improve on the stability of these systems in resisting the onset of loosening, consequently
increasing the longevity of joint replacement systems.
166
8.5 REFERENCES
Australian Orthopaedic Association, T., 2010. National Joint Registry, Hip and Knee
Arthroplasty Annual Report 2010,
http://www.dmac.adelaide.edu.au/aoanjrr/documents/aoanjrrreport_2010.pdf.
Cheung, E. V, O’Driscoll, S.W., 2007. Total elbow prosthesis loosening caused by ulnar
component pistoning. The Journal of Bone and Joint Surgery; American Volume 89,
1269–1274.
Fetterly, S., 2012. Quantification of bone and cement strains surrounding a distal ulanr
implant with varying cement-stem interface conditions. (M.E.Sc Thesis) The
University of Western Ontario, pp. 1-38.
Jeon, I.H., Morrey, B.F., Sanchez-Sotelo, J., 2012. Ulnar component surface finish
influenced the outcome of primary Coonrad-Morrey total elbow arthroplasty.
Journal of Shoulder and Elbow Surgery 21, 1229–1235.
Kedgley, A.E., Takaki, S.E., Lang, P., Dunning, C.E., 2007. The effect of cross-sectional
stem shape on the torsional stability of cemented implant components. Journal of
Biomechanical Engineering 129, 310–314.
Mann, K.A., Bartel, D.L., Wright, T.M., Ingraffea, a R., 1991. Mechanical characteristics
of the stem-cement interface. Journal of Orthopaedic Research 9, 798–808.
New Zealand Orthopaedic Association, T.N.Z.J.R., 2010. Eleven Year Report- January
1999 to December 2009., http://www.cdhb.govt.nz/njr/reports/A2D65CA3.pdf.
167
APPENDICES
168
Appendix A: Thesis Glossary
This appendix defines the terminology used throughout this thesis, to assist with
understanding the technical terms that may not be familiar to the reader.5
Abduct: To draw away from a position near the middle axis of the body (as in a limb).
Alloy: A substance composed of two or more metals, or of a metal and non-metal fused
together when molten.
Attenuate: The decrease or lessening of the amount of energy.
Bioactive: Having the effect of a living organism.
Biocompatibility: The condition of being compatible with living tissue, or not being
toxic to living tissue.
Camera/Image sensor: A device that responds to the stimulus of light and transmits a
resulting impulse.
Cartilage: Flexible connective tissue found in many areas in the bodies of humans and
other animals
Cartilaginous joints: Joints that are connected by cartilaginous material.
Charged couple device (CCD): A semiconductor device, used as an optical sensor,
which stores charge and subsequently transfers it to an amplifier and detector.
Diffraction: A modification that light undergoes, especially in passing by the edges of
opaque bodies or through narrow openings.
5 The medical definitions listed here were obtained from the Marriam-Webster, Medline Plus Medical
Dictionary, a service of the U.S. National Library of Medicine and National Institutes of Health. The
general terminology was modified from the Marriam-Webster online dictionary, and Wikipedia Online
Encyclopdia.
169
Extension: The unbending movement of a joint in a limb, that increases the angle
between the bones of the limb and the joint.
Fibrous joints: Joints that are connected by fibrous, connective tissue such as collagen.
Field-of-view (FOV): The maximum observable area (as seen from a camera).
Flexion: A bending movement around a joint in a limb that decreases the angle between
the bones of the limb and the joint.
f-number: Quantitative measure of the lens speed, and is the ratio of the focal length to
the diameter of the lens opening.
Focal length: Measure of how strongly the system converges or diverges light; the
distance (from the lens) over which the light rays are brought into focus.
Glenohumeral: The connection between the glenoid cavity and humerus of the
shoulder.
Glenoid cavity: The shallow cavity of the upper part of the scapula, in which the
humerus articulates.
Hounsfield unit (HU): Quantitative scale for describing radiodensity of a material,
relative to the radiodensity of distilled water and air at standard pressure and temperature.
Humerus: The longest bone of the upper arm or forelimb extending from the shoulder to
the elbow.
Image distortion: Deviation of projected image, in which straight lines of an object do
not maintain similar straightness in the image.
Implant fixation: Surgical implementation and anchoring of joint replacement systems.
Intramedullary: Situated or occurring within the medulla of bone; use of the marrow
space of a bone for support.
In-vitro: Outside the living body and in an artificial environment.
170
In-vivo: Inside the living body, or within a living organism
Lens aperture: The opening or hole of a lens through which light enters.
Micromotion: Motion, such as that of an implant, occurring on the order of
micrometers.
Monomer: A molecule that may bind chemically to other molecules to form polymers.
Osseointegration: The firm anchoring of a surgical implant by growth of bone around
it, without fibrous tissue formation at the interface
Osteoarthritis: Degenerative joint disease that causes changes in the bone and cartilage
at the joint, resulting in wearing down of joint surfaces.
Osteoporosis: Decrease in bone mass with decreased density and enlargement of bone
spaces.
Perspective error: Difference in the image magnification as a result of object
position/distance from the lens.
Photons: The unit intensity of light; smallest physical quantity of electromagnetic
radiation.
Pixels: Small discrete elements that make up a image; picture elements.
Plasma spray: Coating in which melted or heated materials are sprayed onto a surface.
Polymer: Chemical compound consisting of repeating structural units or molecules.
Stereophotogrammetry: Technique used for the assessment of three-dimensional
migration of joint replacement from bone, using an imaging technique to obtain real-time
moving images of markers attached to the bone and implant.
Resolution: The detail and image holds; quantification of how close lines can be and
still visibly resolved.
171
Sintered bead: Coating in which beads are diffused onto a surface at high temperatures.
Substrate material: Material on which a process is conducted; surface to which coating
is applied.
Synovial joints: Most common of the moveable joints characterized by two bony
surfaces covered with cartilage, contained in a fibrous capsule containing joint fluid.
Telecentric lens: Compound lens for which chief rays are parallel to the optical axis in a
object/image space, making the object appear to be the same size independent of its
location in space.
Tibia: The large, inner bone of the lower leg, located between the knee and ankle.
Ulna: One of the two bones located in the forearm, situated on the side of the forearm
with the little finger, and articulates with the humerus in the elbow joint.
Ulnohumeral: The connection between the ulna and humerus bone of the elbow joint.
172
Appendix B: Specifications for Hardware of Optical System
This appendix details the specifications for the hardware used in the optical system
discussed in Chapters 2,3,4,5 and 6.
Figure B.1: Specifications for the Basler Pilot AG piA 2400- 12gc Camera
(Adapted from the User Manual for GigE Vision Cameras; Document No.AW000151,
Version 16; Basler Vision Technologies)
173
Figure B.2: Specifications for the Opto Engineering Telecentric Lens
(Product test report provided with the telecentric lens)
174
Figure B.3: Specifications for the Advanced Illumination, Axial Diffuse Illumintor
(Adapted from sales sheet for Advanced Illumination products;
http://www.1stvision.com/lighting/AI/uploads/products/DL072.pdf)
175
Appendix C: LabVIEW®
Programs
This appendix shows the LabVIEW® back panels for the custom written programs used
for calibration of the optical system and marker tracking data collection. These
programs were used for Chapters 2 ,3, 4, 5 and 6.
Figure C.1: Back Panel of Calibration Program
The calibration program determines the vertical and horizontal pixel count for a
rectangular region of interest of the calibration grid, set by the user. Based on the known
spacing of the calibration grid (i.e., 0.1 mm), the pixel to millimeter conversion for the
region of interest is found.
176
Figure C.2: Back Panel of the Optical Tracking Program
The optical tracking program incorporating a thresholding method to detect markers
placed on landmarks of interest. The program determines the centroids of the detected
markers, and tracks their respective (x,y) coordinates throughout the duration of the
program. The program outputs the individual (x,y) coordinates of the marker, the length
between the two points, and the corresponding load and position data from the materials
testing machine.
177
Appendix D: Bland-Altman Plots for Optical System Validation
This appendix shows the individual Bland-Altman plots for the various regions of the
optical system field of view (Region 1-9), as discussed in Chapter 2.
Figure D.1: Bland Altman plots for Region 1-9 of Horizontal Displacement
Individual plots for horizontal displacements within the nine regions of the image field of
view. Regions 7 and 9 show some scatter points outside 95% limit of agreement (as
indicated by dashed lines).
178
Figure D.2: Bland Altman plots for Region 1-9 of Vertical Displacement
Individual plots for vertical displacements within the nine regions of the image field of
view. All regions show some scatter points outside 95% limit of agreement (as indicated
by dashed line) for measurements of 0.5 mm.
179
Appendix E: Calculation of Optical System Accuracy
This appendix details the calculations used for determining the accuracy of the optical
system as described in Chapter 2.
Accuracy as measured by percent error for the range of applied displacements (0.005-
0.500 mm).
Procedure 6:
1) Determine line of best-fit for data from optical system (i.e., measured) and
micrometer (i.e., true) (Figure E.1).
Equation of best-fit: (Eq. F.1)
2) Evaluate errors in the measurements by creating a deviation plot from the data sets
(Figure E.2).
( )
( )
3) Accuracy of the system, presented as % error of the range of output displacements:
- Range of output displacements from equation of best-fit:
o 0.507 mm-0.007 mm= 0.500 mm
6 As described in: Wheeler AJ, Ganji AR. 2010. Chapter 2: General Characteristics of Measurement
Systems. In: Introduction to Engineering Experimentation; Pearson Higher Education, NJ. pp 21-24
180
- Positive limit:
( )
- Negative limit:
( )
Therefore, the accuracy of measurements from the optical system is ( ,
) of the measured value.
181
Figure E.1: Line of Best-Fit for Measured against True Displacements
The plot shows the scatter of points, along with the line of best fit relating the measured
data from the optical system to the true data from the micrometer.
182
Figure E.2: Deviation Plot for Optical System Measurements
This deviation plot shows the difference in measurements from the optical system and
micrometer, against true displacements (i.e., micrometer). The dashed lines parallel to
the horizontal axis are the accuracy limits of the data (+0.039 mm, -0.041 mm), such that
data points not considered to be outliers are contained within them.
183
Appendix F: Calculation of Optical System Resolution
This appendix details the calculations used for determining the resolution of the optical
system, based on the calibration method described in Chapter 2.
Calibration Method:
1) The calibration grid (Figure F.1) (grid spacing 0.100 ± 0.002 mm (Pyser-SGI Ltd,
Kent, UK)) was placed in the camera field of view, at the location expected for
marker placement.
2) Using the LabVIEW calibration program described in Appendix C.1, the horizontal
and vertical pixel length (i.e., pixel to millimetre conversion) was determined for the
nine regions of the image field of view.
3) Two trials were done in each region of the image field of view, and an average pixel
size was determined from the measured vertical and horizontal lengths of the pixels.
184
Figure F.1: Calibration Grid used for Pixel to mm Conversion
The calibration grid, with grid spacing 0.100 ± 0.002 mm, was divided into 9 regions of
the image field of view to determine the pixel to mm conversion for each region.
185
Table F.1: Vertical and Horizontal Pixel Lengths
1 1 0.0033 0.0034
1 2 0.0033 0.0035
1 3 0.0033 0.0033
1 4 0.0034 0.0034
1 5 0.0033 0.0033
1 6 0.0033 0.0033
1 7 0.0033 0.0034
1 8 0.0034 0.0034
1 9 0.0033 0.0033
2 1 0.0033 0.0034
2 2 0.0033 0.0034
2 3 0.0033 0.0034
2 4 0.0034 0.0033
2 5 0.0034 0.0033
2 6 0.0034 0.0033
2 7 0.0033 0.0033
2 8 0.0033 0.0033
2 9 0.0033 0.0034
Average pixel size 0.003 0.003
Trial # Region FOV Pixel vertical length (mm) Pixel horizontal length (mm)
186
Uncertainty in calibration measurement 7,8
:
(
)
(Eq. G.1)
Where:
UC = Total uncertainty
BC = Systematic error, based on manufacturer specifications of grid
PC = Random error in repeated measurements of pixel size
Random Error in Repeated Measurements of Pixel Size:
(Eq. G.2)
Where:
= standard deviation of the mean from repeated pixel to mm conversions
(n=36)
t = t-statistic with 95% confidence interval for 36 repeated pixel to mm
conversions, as obtained from the student’s t-distribution table
(
√ ) (Eq. G.3)
7 Wheeler AJ, Ganji AR. 2010. Chapter 7: Experimental Uncertainty Analysis. In: Introduction to
Engineering Experimentation (3rd
Ed); Pearson Higher Education, NJ. pp 199-230
8 Figliola RS, Beasley DE.2006. Chapter 5: Uncertainty Analysis. In: Theory and Design for Mechanical
Measurements (4th
Ed); John Wiley and Sons, NJ. pp 149-182
187
Therefore, total uncertainty for pixel to mm conversion from equation G.1:
(
)
( ( (
√ ))
)
Optical System Resolution = 0.003 ± 0.002 mm
188
Appendix G: Engineering Drawings and Assemblies
This appendix includes the dimensioned drawing of parts and assemblies for the stems
and jigs used in Chapters 3,4,5,6 and 7 of this thesis. The models were created in
Solidworks (Dessault Systems, Concord, MA).
Figure G.1: Engineering Drawing of Smooth Implant Stem
189
Figure G.2: Engineering Drawing of 20 mm Length Plasma Spray Implant Stem
190
Figure G.3: Engineering Drawing of 20 mm Length Beaded Implant Stem
191
Figure G.4: Engineering Drawing of 10 mm Length Plasma Spray Implant Stem
192
Figure G.5: Engineering Drawing of 10 mm length Beaded Implant Stem
193
Figure G.6: Engineering Drawing of 0.6 mm Grooved Implant Stem
194
Figure G.7: Engineering Drawing of 1.1 mm Grooved Implant Stem
195
Figure G.8: Engineering Drawing of Parts for Potting Jig Assembly
196
Figure G.9: Engineering Drawing of Delrin® Block used to Centralize Stems for
Potting
197
Figure G.10: Engineering Assembly of Jig used for securing Stem in Torsion and
Compression
(Assembly drawing done by University Machine Services)
198
Figure G.11: Engineering Drawing of Securing Plates used for Bending Tests
199
Figure G.12: Engineering Assembly of Cement Sample Template
200
Figure G.13: Engineering Drawing of Parts for the Cement Sample Template
201
Figure G.14: Engineering Assembly of Stem Holder used in µCT-imaging Bend
Tests
202
Figure G.15: Engineering Drawing of Parts for Stem Holder Assembly used in
µCT-imaging Bend Tests
203
Figure G.16: Engineering Drawing of Back Plate for Stem Holder Assembly used in
µCT-imaging Bend Tests
204
Appendix H: Cement Preparation and Potting Technique
This appendix describes the cementing procedure used in Chapters 3, 4, 5 and 6 of this
thesis, for fixation of the implant stems. The technique details the centralization and
positioning of the implant stem (Appendix G) for potting, the mixing of bone cement, and
application of cement to the aluminum tubes containing the stems.
Preparation of Stem and Cement
In preparation for cementing the following are needed: implant stems, aluminum tubes,
delrin blocks, adhesive tape, potting jig.
1) Prior to cementing, clean all stems and aluminum tubes with Acetone, and rinse
with distilled water.
2) Insert Delrin® blocks (19 mm x 19 mm x 30 mm) (Appendix G.9) into the base of
the aluminum tubes (internal diameter: 19 mm x 19 mm x 50 mm), and secure the
tube’s base using adhesive tape. These blocks are machined with an extruded
through-cut down their center (diameter 8mm), through which stems will be
positioned along the central axis of the aluminum tubes.
3) Insert stems into the Delrin® blocks and carefully position, ensuring all stem are
aligned along the central axis of the tubes. These Delrin®
blocks will also act as
stoppers to control the depth to with the stems are potted in subsequent steps of
this process (Figure H.1).
4) Once the stems are appropriately positioned within the aluminum tubes, secure all
tubes within the potting jig for application of bone cement (Figure H.1). This jig
will allow the maintenance of stem alignment during the cementing and curing
process, as well as eliminate motion of the stem and aluminum tube during the
cement application.
205
Figure H.1: Potting Jig used for Securing Stems during Cementing
(A) All stems are centralized into aluminum tubes using Delrin®
blocks placed at the base
of the tube (indicated by black dashed double arrow). The blocks ensure appropriate
positioning of the stem along the central axis of the aluminum tube, as well as control the
depth to which the stems are potted (20 mm). (B) Once stems are aligned, the aluminum
tubes with stems are placed in the potting jig for cement application and curing.
206
Mixing and Application of Bone Cement
1) Place the package of Simplex P® bone cement in freezer at least 24 hrs prior to
cement preparation, to allow for chilling of the monomer liquid and polymer
powder.
2) Within the fume hood of the laboratory, attach the vacuum mixing system
(Optivac©
, Biomet Inc., Warsaw, Indiana, USA) to a vacuum pump maintaining
the pressure at 15-20 mmHg. The temperature within the fume hood should be
maintained at 22 ºC.
3) Place the provided funnel over the larger cartridge (N.B., a small and large
cartridge is included in the vacuum mixing system), and empty one packet of the
polymer powder into the cartridge.
4) Remove the funnel, and empty the monomer liquid into the cartridge already
containing the powder.
5) Immediately close the vacuum container, and allow 10 seconds for the vacuum
pressure to form within the system.
6) Begin rapid plunging, rotating the plunger through the cement during each
plunging motion. Ensure that the plunger makes contact with both ends of the
cartridge to allow the polymer powder to fully integrate with the monomer liquid.
7) After 60 seconds of plunging, the cement mixture should have a viscous
consistency. Pour the entire mixture into a 60 ml syringe, and immediately insert
the syringe’s plunger.
8) Tilt the syringe upwards and slowly plunge the mixture towards the application
tip, to remove the excess air within the syringe.
207
9) Use the syringe to apply the cement mixture to the aluminum tubes containing the
implant stems (Figure H.1).
10) Leave the cemented construct within the fume hood for 24 hours, at the
controlled 22 ºC, for curing.
208
Appendix I: Tabulated Data
Table I.1: Tabulated Data for Titanium Surface Treated Implant Stems in Torsion
TITANIUM
Surface Treatment Stem # Torque (Nm) Toggle Rotation (deg) Normalized Toggle (deg/Nm)
Smooth 1 10.00 0.650 0.065
2 10.00 0.530 0.053
3 5.00 0.260 0.052
4 12.00 0.410 0.034
5 4.00 0.310 0.078
Average 8.20 0.432 0.056
SD 3.49 0.160 0.016
Beaded 20mm 1 30.00 0.400 0.013
2 30.00 0.570 0.019
3 30.00 0.650 0.022
4 30.00 0.380 0.013
5 30.00 0.780 0.026
Average 30.00 0.556 0.019
SD 0.00 0.169 0.006Plasma Spray 20mm 1 30.00 1.070 0.036
2 30.00 1.610 0.054
3 30.00 0.840 0.028
4 30.00 0.790 0.026
5 30.00 1.290 0.043
Average 30.00 1.120 0.037
SD 0.00 0.339 0.011
209
Table I.2: Tabulated Data for Cobalt Chrome Surface Treated Implant Stems in
Torsion
COBALT CHROME
Surface Treatment Stem # Torque (Nm) Toggle Rotation (deg) Normalized Toggle (deg/Nm)
Smooth 1 9.00 0.320 0.036
2 18.00 0.200 0.011
3 13.00 0.450 0.035
4 13.00 0.310 0.024
5 5.00 0.270 0.054
Average 11.60 0.310 0.032
SD 4.88 0.091 0.016
Beaded 20mm 1 16.00 0.590 0.037
2 16.00 0.420 0.026
3 17.00 0.510 0.030
4 16.00 0.540 0.034
5 17.00 0.650 0.038
Average 16.40 0.542 0.033
SD 0.55 0.086 0.005
Plasma Spray 20mm 1 30.00 0.700 0.023
2 30.00 0.600 0.020
3 30.00 0.530 0.018
4 30.00 0.430 0.014
5 30.00 0.730 0.024
Average 30.00 0.598 0.020
SD 0.00 0.123 0.004
210
Table I.3: Tabulated Data for Titanium Surface Treated Implant Stems under
Compression
TITANIUM
Smooth 1 0.025 - 3.00 202 4.13E-05 -
2 0.009 - 3.00 201 1.49E-05 -
3 0.006 - 3.00 201 9.95E-06 -
4 0.031 - 3.00 206 5.02E-05 -
5 0.005 - 2.00 101 2.48E-05 -
6 0.028 - 3.00 201 4.64E-05 -
Average 0.017 - 2.83 185 3.12E-05 -
SD 0.012 - 0.41 41 1.70E-05 -
Bead 20mm 1 0.160 0.149 10.00 25900 6.18E-07 5.96E-06
2 0.050 0.207 10.00 25900 1.93E-07 8.28E-06
3 0.120 0.211 10.00 25900 4.63E-07 8.44E-06
4 0.040 0.179 10.00 25900 1.54E-07 7.16E-06
5 0.037 0.194 10.00 25900 1.43E-07 7.76E-06
6 0.080 0.207 10.00 25900 3.09E-07 8.28E-06
Average 0.081 0.191 10.00 25900 3.13E-07 7.65E-06
SD 0.050 0.024 0.00 0 1.92E-07 9.51E-07
Bead 10mm 1 0.121 0.635 10.00 1889 6.41E-06 6.42E-04
2 0.080 1.214 10.00 4583 1.75E-06 3.30E-04
3 0.072 0.339 10.00 2269 3.17E-06 2.48E-04
4 0.100 0.417 10.00 10435 9.58E-07 4.37E-05
5 0.090 0.923 10.00 4499 2.00E-06 2.56E-04
6 0.100 1.156 10.00 4091 2.44E-06 3.62E-04
Average 0.094 0.781 10.00 4628 2.79E-06 3.14E-04
SD 0.017 0.373 0.00 3069 1.92E-06 1.95E-04
Plasma Spray 20mm 1 0.120 0.800 10.00 2349 5.11E-06 5.52E-04
2 0.163 0.851 10.00 1980 8.23E-06 7.88E-04
3 0.100 1.120 10.00 9469 1.06E-06 1.31E-04
4 0.163 0.989 10.00 1601 1.02E-05 1.41E-03
5 0.108 0.751 10.00 1288 8.39E-06 1.94E-03
6 0.102 0.872 10.00 1610 6.34E-06 1.23E-03
Average 0.126 0.897 10.00 3050 6.55E-06 1.01E-03
SD 0.029 0.135 0.00 3166 3.22E-06 6.48E-04
Plasma Spray 10mm 1 0.150 - 8.00 722 2.60E-05 -
2 0.142 - 8.00 791 2.24E-05 -
3 0.141 - 9.00 871 1.80E-05 -
4 0.181 - 9.00 842 2.39E-05 -
5 0.175 - 9.00 887 2.19E-05 -
6 0.115 - 9.00 805 1.59E-05 -
Average 0.151 - 8.67 820 2.13E-05 -
SD 0.024 - 0.52 60 3.76E-06 -
NormalizedToggle
(mm/kN*#cycles)
Normalized Stem
Motion Post-10kN Surface Treatment Stem # Interface Toggle (mm)
Stem Motion Post-
10kN (mm)
Load
(kN)
#
Cycles
211
Table I.4: Tabulated Data for Cobalt Chrome Surface Treated Implant Stems
under Compression
COBALT CHROME
Smooth 1 0.003 - 2.00 101 1.49E-05 -
2 0.020 - 3.00 201 3.25E-05 -
3 0.007 - 3.00 232 1.01E-05 -
4 0.018 - 3.00 201 2.99E-05 -
5 0.006 - 2.00 101 2.97E-05 -
6 0.007 - 2.00 101 3.47E-05 -
Average 0.010 - 2.50 156 2.53E-05 -
SD 0.007 - 0.55 61 1.02E-05 -
Beaded 20mm 1 0.100 0.195 10.00 25900 3.86E-07 7.80E-06
2 0.080 0.201 10.00 25900 3.09E-07 8.04E-06
3 0.045 0.373 10.00 7965 5.65E-07 5.28E-05
4 0.019 0.201 10.00 25900 7.34E-08 8.04E-06
5 0.018 0.208 10.00 25900 6.95E-08 8.32E-06
6 0.150 0.165 10.00 25900 5.79E-07 6.60E-06
Average 0.069 0.224 10.00 22911 3.30E-07 1.53E-05
SD 0.052 0.075 0.00 7322 2.26E-07 1.84E-05
Beaded 10mm 1 0.018 - 3.00 201 2.97E-05 -
2 0.005 - 3.00 204 8.17E-06 -
3 0.005 - 3.00 209 7.97E-06 -
4 0.006 - 2.00 101 2.72E-05 -
5 0.049 0.415 10.00 25900 1.91E-07 1.66E-05
6 0.011 - 3.00 201 1.82E-05 -
Average 0.016 0.415 4.00 4469 1.52E-05 1.66E-05
SD 0.017 - 2.97 10499 1.18E-05 -
Plasma Spray 20mm 1 0.160 1.045 10.00 6621 2.42E-06 1.83E-04
2 0.165 1.175 10.00 7155 2.31E-06 1.88E-04
3 0.100 0.252 10.00 25900 3.86E-07 1.01E-05
4 0.129 1.206 10.00 15323 8.41E-07 8.36E-05
5 0.133 1.124 10.00 15700 8.47E-07 7.59E-05
6 0.162 1.190 10.00 9990 1.62E-06 1.31E-04
Average 0.141 0.999 10.00 13448 1.40E-06 1.12E-04
SD 0.026 0.370 0.00 7236 8.43E-07 6.87E-05
Plasma Spray 10mm 1 0.100 9.00 804 1.38E-05 -
2 0.193 0.028 10.00 911 2.12E-05 -
3 0.114 0.869 10.00 1139 1.00E-05 3.64E-03
4 0.125 0.148 10.00 923 1.35E-05 -
5 0.159 0.02 10.00 909 1.75E-05 -
6 0.145 9.00 824 1.96E-05 -
Average 0.139 0.266 9.67 918 1.59E-05 3.64E-03
SD 0.034 0.406 0.52 119 4.20E-06 -
Normalized Stem
Motion Post 10kN Interface Toggle (mm)
#
CyclesSurface Treatment Stem #
Load
(kN)
Stem Motion post-
10kN (mm)
Normalized Toggle
(mm/kN*#cycles)
212
Table I.5: Tabulated Data for Circumferential Grooved Implant Stems under
Compression
COBALT CHROME
Smooth 1 3000 201 0.004 6.08E-06 - -
2 3000 201 0.006 9.95E-06 - -
3 3000 201 0.004 7.19E-06 - -
4 3000 201 0.012 2.05E-05 - -
5 3000 201 0.002 3.32E-06 - -
Average 3000 201 0.006 9.40E-06 - -
SD 0 0 0.004 6.62E-06 - -
Grooved 1.1mm 1 10000 25900 0.054 2.10E-07 0.457 1.83E-05
2 10000 25900 0.099 3.84E-07 0.449 1.80E-05
3 10000 25900 0.097 3.76E-07 0.420 1.68E-05
4 10000 25900 0.084 3.23E-07 0.268 1.07E-05
5 10000 25900 0.043 1.66E-07 0.308 1.23E-05
Average 10000 25900 0.076 2.92E-07 0.380 1.52E-05
SD 0 0 0.026 9.87E-08 0.087 3.47E-06
Grooved 0.6mm 1 10000 25900 0.066 2.54E-07 0.133 5.32E-06
2 10000 25900 0.085 3.29E-07 0.295 1.18E-05
3 10000 25900 0.041 1.58E-07 0.320 1.28E-05
4 10000 25900 0.046 1.79E-07 0.253 1.01E-05
5 10000 25900 0.025 9.65E-08 0.248 9.92E-06
Average 10000 25900 0.053 2.03E-07 0.250 9.99E-06
SD 0 0 0.023 9.01E-08 0.072 2.87E-06
Normalized Stem
Motion post 10kN
Stem Motion
post 10KN Surface Treatment Stem #
Load
(N)
#
cycles
Interface
Toggle (mm)
Normalize Toggle
(mm/N)
213
Table I.6: Tabulated Data for Circumferential Grooved Implant Stems in Torsion
COBALT CHROME
Surface Treatment Stem # Torque (Nm) Toggle Rotation (deg) Normalized Toggle (mm/Nm)
Smooth 1 9 0.200 0.022
2 18 0.163 0.009
3 13 0.322 0.025
4 13 0.240 0.018
5 5 0.189 0.038
Average 11.600 0.223 0.022
SD 4.879 0.062 0.010
Grooved 1.1mm 1 17 0.720 0.042
2 19 0.719 0.038
3 15 0.749 0.050
4 18 0.802 0.045
5 15 0.569 0.038
Average 16.800 0.712 0.043
SD 1.789 0.087 0.005
Grooved 0.6mm 1 20 0.194 0.010
2 20 0.459 0.023
3 14 0.422 0.030
4 10 0.224 0.022
5 23 0.450 0.020
Average 17.40 0.350 0.021
SD 5.27 0.130 0.007
214
Table I.7: Tabulated Data for Titanium Surface Treated Implant Stems in Bending
TITANIUM
Surface Treatment Stem # Interface Toggle (mm) Offset Stem Motion (mm)
Smooth 1 0.065 0.045
2 0.066 0.043
3 0.052 0.036
4 0.067 0.050
5 0.083 0.052
Average 0.067 0.045
St Dev 0.011 0.006
Bead 20mm 1 0.050 0.034
2 0.073 0.045
3 0.054 0.034
4 0.040 0.026
5 0.047 0.030
Average 0.053 0.034
St Dev 0.012 0.007
Bead 10mm 1 0.054 0.039
2 0.066 0.036
3 0.046 0.026
4 0.044 0.027
5 0.047 0.032
Average 0.051 0.032
St Dev 0.009 0.006
Plasma Spray 20mm 1 0.064 0.034
2 0.059 0.038
3 0.074 0.042
4 0.058 0.035
5 0.057 0.037
Average 0.062 0.037
St Dev 0.007 0.003
Plasma Spray 10mm 1 0.068 0.048
2 0.085 0.066
3 0.087 0.064
4 0.095 0.066
5 0.051 0.028
Average 0.077 0.054
St Dev 0.018 0.017
215
Table I.8: Tabulated Data for Cobalt Chrome Surface Treated Implant Stems in
Bending
COBALT CHROME
Surface Treatment Stem # Interface Toggle (mm) Offset Stem Motion (mm)
Smooth 1 0.091 0.071
2 0.060 0.049
3 0.131 0.102
4 0.057 0.014
5 0.070 0.051
Average 0.082 0.057
St Dev 0.031 0.032
Bead 20mm 1 0.044 0.027
2 0.056 0.039
3 0.037 0.024
4 0.040 0.031
5 0.033 0.015
Average 0.042 0.027
St Dev 0.009 0.009
Bead 10mm 1 0.042 0.039
2 0.122 0.099
3 0.093 0.087
4 0.037 0.033
5 0.046 0.030
Average 0.068 0.058
St Dev 0.038 0.033
Plasma Spray 20mm 1 0.061 0.042
2 0.051 0.035
3 0.056 0.034
4 0.050 0.025
5 0.040 0.029
Average 0.052 0.033
St Dev 0.008 0.006
Plasma Spray 10mm 1 0.038 0.036
2 0.040 0.025
3 0.061 0.054
4 0.062 0.044
5 0.063 0.044
Average 0.053 0.041
St Dev 0.013 0.011
216
Appendix J: Letter of Permission
This appendix includes the letter of permission from the publisher, for the re-print of
published material within Chapter 4 of this thesis.
217
218
Appendix K: Bead Embedment Procedure
This appendix describes the procedure used in Chapter 7 for embedding steel bead
markers into bone cement during the cementing process. The methodology highlights the
cement mixing and application technique used to ensure appropriate suspension of the
beads within the cement mantle.
1) Using the cement preparation technique described in Appendix H, vacuum mix
the pre-chilled bone cement in the controlled environment of the fumehood.
2) After 60 seconds of vacuum-mixing, pour the mixture from the vacuum cartridge
into an open bowl, and continue mixing with a spatula for an additional 60
seconds.
3) After 60 seconds of hand mixing, the bead-cement mixture should appear doughy
in consistency. At this time-point, incorporate the steel beads into the cement and
continue mixing. With the doughy consistency of the cement, the beads should
remain suspended (as opposed to sinking) within the mixture.
(N.B. If the cement mixture does not appear doughy after 60 seconds, continue hand
mixing until the doughing phase begins. ONLY incorporate the beads into the mixture
when this consistency is reached. This is important for appropriate suspension of the
beads within the cement)
4) When the beads appear to be fully integrated into the cement, pour the mixture
into the 60ml syringe, and immediately insert the syringe’s plunger.
5) Tilt the syringe upwards and slowly plunge the mixture towards the application
tip, to remove excess air within the syringe.
6) Use the syringe to apply the bead-cement mixture to aluminum tubes containing
the stem, or to the cement template used for moulding cement samples.
219
Appendix L: Mechanical Testing of Beaded Cement Samples
This appendix describes the three-point bend tests used to compare cement samples
prepared with different bead-cement ratios. The results from this appendix was used to
determine the appropriate bead-cement ratio to be used in Chapter 7, ensuring
mechanical integrity of the bone cement, while allowing optimum dispersion of beads
within the cement mantle.
Introduction:
The incorporation of steel beads into bone cement can be useful as imaging
markers, when studying the internal mechanics of the cemented systems using CT-
imaging techniques. The vast difference in density values between the steel beads and
the surrounding bone cement make the beads appropriate as high contrast markers within
the cement. However, this difference in density values may also affect the mechanical
properties of bone cement. As such, this pilot study aimed to determine the change in
mechanical properties of bone cement by incorporation of various bead-cement ratios.
Therefore, the purpose of this study was to investigate the effect of different bead-cement
ratios on the flexural modulus of Simplex P® bone cement.
Materials and Methods:
Cement preparation was done using the procedure described in Appendix I, where
the beads were incorporated into the bone cement during the doughing phase of cement
preparation. Four ratios of cement-bead mixtures were made using four separate
packages of pre-chilled Simplex P® bone cement; 0%, 0.15%, 0.2%, and 0.25%. Bead
ratios represented the volume of beads relative to the volume of bone cement. One
package of cement powder mixed with monomer liquid created a cement mixture volume
of approximately 40 cm3.
220
Once the beads appeared fully integrated into the cement, the mixture was poured
into the syringe for subsequent application to an aluminum cement template. This
template created five cement samples of dimensions of 80 mm x 10 mm x 3.3 mm, based
on specifications from the ISO 5833 standards9 (Figure L.1). The cement samples were
left to cure for 24 hrs within the template, after which they were removed for mechanical
testing.
The remaining bead-cement mixture that was left over from creation of cement
samples was incorporated into aluminum tubes containing a stem (Appendix I), and
potted to a fixed depth of 10 mm. These cement mantles were used for micro-CT
imaging to determine the internal distribution of the beads within the aluminum tubes.
Using this bead distribution, along with the results of mechanical testing (see below), the
appropriate bead ratio was determined for embedment into the cement mantles.
Mechanical testing was done using a uni-axial materials testing machine (Instron
8872, Canton, MA). The cement samples were positioned on a three-point bending
platform placed at the base of the materials testing machine (Figure L.2), and exposed to
monotonic loading at a rate of 2 mm/min. The samples were tested to failure, defined by
the sudden decrease in load after initial linear increase, accompanied by instantaneous
fracture of the cement samples.
The slope of the force-displacement graphs, generated by the data from the
materials testing machine, was used to calculate the bending modulus of the 0%, 0.15%,
0.2% and 0.25% cement samples. The respective moduli were compared to determine
the effect of bead ratio on the mechanical properties of bone cement.
9 ISO International Standards, ISO 5833/1. Implants for Surgery. Acrylic Resin Cements. Part I:
Orthopaedic Applications.
221
Figure L.1: Bead Embedded Cement Samples
Cement samples used to determine the effect of bead ratio on the mechanical properties
of bone cement. From top to bottom; 0%, 0.15%, 0.20%, 0.25%.
222
Figure L.2: Experimental Set-up for Three-point Bending of Cement Samples
The cement samples were placed on the three-point bending platform for application of
loads at a rate of 2 mm/min. Failure was defined as a sudden decrease in load
accompanied by instantaneous fracture of the cement sample.
223
Table L.1: Bending Modulus of Bone Cement with Various Bead- Cement Ratios
224
Results:
From visual analysis of the five samples created for each bead ratio, it was
observed that the 0.25% samples showed the greatest bead distribution compared to the
other bead ratios. In comparison, the 0.15% beaded sample showed the least bead
distribution.
Results of mechanical testing showed that the addition of beads caused an overall
increase in the bending modulus of bone cement, as seen in Table K.1. In addition, it was
found that the 0.15% sample showed the greatest variability in its mechanical response to
the three-point bend tests, based on its large standard deviation, compared to the other
cement samples.
CT imaging of the various cement mantles found that the 0.20% and 0.25% bead
ratios showed suitable dispersion of beads within the samples of cement, with the 0.15%
sample showing most inconsistent distribution.
Discussion:
The cementing technique previously described in Appendix J, proved useful for
embedment of steel beads within the bone cement. From the observations of bead
distribution within the cement samples, all beads appeared reasonably dispersed
throughout the sample volumes, with the 0.25% bead ratio showing greatest dispersion
and bead consistency within its samples (Figure L.1).
Mechanical testing showed that the addition of beads to bone cement increased
the stiffness of the cement samples. When comparing the bending modulus of bone
cement observed in this study, to that found within the literature, it was seen that the
225
modulus for the 0% bead sample (2.89 GPa) was comparable to that described by Lee 10
(2.55 GPa).
Imaging analysis found that the 0.20% and 0.25% bead ratios showed satisfactory
dispersion of beads within the cement mantle surrounding the stem, and it was
determined that either of the ratios would be optimal for bead visualization in Chapter 7.
When considering the appropriate bead ratio for inclusion in CT analysis of the
cemented construct, it was determined that although 0.25% sample showed the greatest
bead dispersion, it caused the largest change in the mechanical properties of bone cement.
In comparison, the 0.15% sample showed the least change in the mechanical response of
bone cement, but the variability among its measurements was quite high. In addition, CT
analysis of its cement mantle showed minimal appearance and distribution of beads. As
such, it was decided that the 0.2% bead ratio would be optimum for the use as cement
markers in subsequent CT imaging studies.
10 Lee, C. 2005, “Properties of Bone Cement: The Mechanical Properties of PMMA Bone Cement” In:
Breusch, S.J., Malchau, H (eds), The Well-Cemented Total Hip Arthroplasty, Springer Verlag, Germany,
pp 61.
226
Appendix M: Custom-built Loading Device
This appendix details the custom built loading device used for testing implant stems in
Chapter 7 of this thesis. The device was initially developed by a fourth year design group
(Goutam Datta, Micheal Dottor, Emily Harvey, Geoffrey McLellan) 11
in the Department
of Mechanical and Materials Engineering at Western University, under the supervision of
Dr. Cynthia Dunning and Dr. Jeffrey Wood, and modified slightly for use in this thesis.
The initial design of the loading device was developed based on the requirements
for a CT-compatible, displacement controlled, loading system, which could apply static
and cyclic loads to the human ulna, while simultaneously collecting real-time image data.
Based on the specific environment required for the loading device (i.e., within the bore of
a micro-CT scanner, while x-ray images are acquired), the design constraints were
focused on the CT compatibility, physical size requirements and load requirements.
CT Compatibility
The device was intended to be place within the CT scanner to acquire real-time
images of the loaded specimen. As a result, the material specifications were limited to
polymers and low density metals, which would limit artifacts in the resultant CT images.
Therefore, the final design incorporated aluminum metal for the fixtures and loading
platform, with Delrin® and carbon fibre for the load applicator and actuating parts of the
device.
Size Requirement
In addition to the material constraints imposed by the CT environment, the
dimensions of the device were limited to the size of the micro-CT scanner’s bore. The
scanner’s bore is 15 cm in diameter, which restricted the design and configuration of the
device to this height and width. As such, all parts were machined to these constraints,
11 Datta, G., Dottor M., Harvey E., McLellan G., 2007 Cyclic Loading Device for Micro CT Scanner,
Department of Mechanical and Materials Engineering, The University of Western Ontario.
227
and the mechanical parts of the device were chosen to function within the restricted
volume.
Load Requirements
The loading requirements for the initial design were based on the load expected to
test the material properties of human ulna. As such, the device was designed to apply
maximum loads of 100 N, as well as maintain rigidity under these loads. For the
purposes of the implant stem testing within this thesis, however, the loading requirements
were increased to 1000 N. Therefore minor modifications were made by replacing the
load cell, and motor used for actuation.
Final Design of Loading Device
The mechanical parts of the loading device consist of a stepper motor connected
to an offset cam, which is further attached to a carbon fibre beam (Figure M.1). The
motor provides rotational motion, which is converted to vertical translation of the carbon
fibre beam through the offset bearing cam. Translation is recorded by a linear variable
differential transducer (LVDT) located near cam. Vertical translation of the pivoted
carbon fibre beam produces a see-saw motion along the length of the beam.
The entire mechanical set-up is housed on an aluminum platform, which is
designed to be attached the bed of the CT-scanner. An aluminum holding jig further
attached to this platform secures the specimen in a horizontal orientation above the
carbon fibre beam (Figure M.2). Therefore, during one rotation of the motor, the see-saw
motion of the carbon fibre beam applies a perpendicular load to the head of the specimen.
This load is measured by a load cell positioned just below the specimen.
228
Figure M.1: Mechanical Parts of Loading Device
The motor connected to the offset cam, which was further connected to a pivoted carbon
fibre beam. During one rotation of the motor, the offset cam facilitated vertical
translation of the pivoted carbon fibre beam, resulting in a see-saw motion along the
length of the beam
229
Figure M.2: Custom-Built Loading Device
The mechanical parts were attached to the loading platform, which was fixated to the bed
of the CT scanner. A holding jig secured the specimen in a horizontal orientation above
the carbon fibre beam, allowing cantilever bending of specimen during one rotation of the
motor.
230
Appendix N: Surface Deviation Analysis
This appendix describes the process used to analyze stem surface deviations between the
unloaded and loaded condition in Chapter 7 of this thesis. This process allowed
comparison of stem-cement motion during the application of bending loads.
Immediately after the scanning protocol is complete, the image data from CT-
scanner is reconstructed into .vff files to be viewed as a 3-D volume made up individual
image slices. This reconstruction is done within MicroView 3D Image Viewer and
Analysis Tool (Parallax Innovations, Ilderton, ON). Once the files are reconstructed, they
are further converted to surface files (.stl) within MicroView, to be exported to, and
analyzed with Geomagic® Qualify (Geomagic, Morrisville, NC). The process below
describes the series of steps used for analysis of the CT data files; from the conversion of
the reconstructed images (.vff) into surfaces (.stl), to the analysis tools used within
Geomagic® Qualify.
STEP 1: Conversion of .vff’s to .stl’s in MicroView
1) Open the .vff image file within MicroView, File Open Select file. The file
should contain a series of image slices which make up the volume of the imaged
object. By placing the mouse cursor over the image volume and left clicking
while dragging the mouse, the orientation of the image volume can be
manipulated.
2) Within the MicroView program, the ‘Tools & Applications’ tab should be
located on the left side of the window. Click on the ‘Standard ROI’ icon under
the Tools & Applications tab. This will introduce a yellow region of interest
(ROI) box in the image volume on the right.
231
3) Within the Standard ROI tab, the ‘Box’ tool and ‘Millimeter’ unit is set as the
default options. These defaults may remain as is.
4) To select the region of interest from which the surface files will be created, place
the mouse cursor over each of the four walls of the yellow ROI box, and click on
the mouse’s wheel to extend these walls in the horizontal and vertical directions
over the desired region. This will cause a change in the ROI size and ROI center
(x,y,z) coordinates shown on the tab located to the left of the program window.
The ‘ROI size’ and ‘ROI center’ coordinates can be altered to make fine
adjustments to the region of interest.
5) Once the desired ROI is shown within the yellow ROI box, select the ‘Visualize’
tab on the toolbar located at the top of the program window, and choose the
‘Isosurface’ option.
(Within the Isosurface tab, various isosurfaces can be created dependent on the image
threshold chosen for the ROI. For analysis of the cemented implant stems, two different
isosurfaces are created; one of the cemented stem contained in the aluminum tube
(Figure N.1A), and the second of the stem and embedded cement beads without the
aluminum tube (Figure N.1B). An image threshold value of 2000 is typically chosen for
the isosurface with the aluminum tube, while the isosurface of the implant stem with
beads is set to an image threshold of 3000)
6) Under the ‘Isosurface Properties’ of the ‘Isosurface tab’, insert the desired
image threshold value (see above) required for creation of the isosurface. Adjust
the surface quality factor to 0.75, and click the ‘Update’ icon. This will create the
desired surface from the image ROI on the right. Once the surface is deemed
suitable, select the ‘Save Surface’ option, and save the resultant isosurface as a
.stl file to the desired location (Figure N.1).
232
Figure N.1: Region of Interest (ROI) used to Create Isosurfaces
(A) The region of interest used to create isosurface for the stem with aluminum tube, and
(B) the same region of interest updated with a new image threshold value, to create
isosurface of the stem and beads only.
233
STEP 2: Analysis using Geomagic Qualify®
Within Geomagic, a single study compares the loaded and unloaded condition for
each of the implant stems. This is done by first opening the isosurface of unloaded stem,
and setting this as the reference surface. Subsequently, the isosurface of the loaded stem
can be imported and set as the test surface. Analysis tools within Geomagic can then be
used to compare the deviation of the test surface (loaded) relative to the reference surface
(unloaded) (Figure N.2).
In order compare the isosurfaces, however, the images must be registered to
ensure that stem motion is evaluated within the same coordinate system. As such, the
isosurfaces of the cemented stem within aluminum tubes are used for registration, by
matching the faces of the reference and test aluminum tubes to one another. This allows
comparison of stem motion within the reference frame of the aluminum tubes.
Step 2a: Registration of Aluminum Tubes in Geomagic
1) Open the .stl containing the isosurface of the unloaded stem contained within the
aluminum tube: File Open Select file
2) The file should appear within the ‘Model Manager’ tab to the left of the program
window. Right click on the file, and select the option “Set as Reference”. This
will result in the abbreviation ‘REF-’ appearing adjacent to the file name in the
Model Manager window.
3) Import the .stl containing the isosurface of the loaded stem contained within the
aluminum tube: File Import Select file
4) Right click the imported file in the Model Manager tab, and select the option “Set
as Test”. This will result in the abbreviation ‘TEST-’ appearing adjacent to the
file name in the Model Manager window.
234
5) Select the rectangular icon (6th
icon from the top) located on the vertical toolbar
at the right of the program window. Use this option to highlight the rectangular
faces of the aluminum tubes by outlining the edges of the faces with the
rectangular cursor.
6) Once all faces of the aluminum tubes are highlighted for the ‘REF-’ and ‘TEST-’
surfaces, click on the ‘Home’ tab located at the top of the program window, and
select the ‘Best Fit’ option. By applying the best fit alignment tool, all surfaces
of the aluminum tubes will be registered.
7) To save the transformation matrix used to align the aluminum tubes, click on the
‘Tools’ tab located at the top of the program window, and select the ‘Transform’
option. This will display the individual translation and rotation matrices that
make up the transformation matrix. Choose the ‘Save Matrix’ option below these
coordinates, and save the matrix to the desired folder.
The transformation matrix saved from the registration of the aluminum tube surfaces will
be used to align the isosurfaces of the implant stem and beads without the aluminum tube.
235
Step 2b: Analysis of Surface Deviation Between the Unloaded and Loaded Stem
1) Open the .stl containing the isosurface of the unloaded implant stem and beads
without the aluminum tube: File Open Select file
2) The file should appear within the ‘Model Manager’ tab to the left of the program
window. Right click on the file, and select the option “Set as Reference”. This
will result in the abbreviation ‘REF-’ appearing adjacent to the file name in the
Model Manager window.
3) Import the .stl containing the isosurface of the loaded implant stem and beads
without the aluminum tube: File Import Select file
4) Right click the imported file in the Model Manager tab, and select the option “Set
as Test”. This will result in the abbreviation ‘TEST-’ appearing adjacent to the
file name in the Model Manager window.
5) Click on the ‘Alignment’ tab located at the top of the program window. Select
the ‘Load Matrix’ option. Choose the transformation file saved in (7) of Step 2a.
This will align the unloaded and loaded surfaces of stem and beads relative to one
another, based on the transformation matrix used to register the aluminum tubes.
6) To compare the surface deviation between the the loaded to unloaded isosurfaces,
click the ‘Home’ tab at the top of the program window, and select the ‘3D
Compare’. This will create a 3D- color plot comparing the deviation in the stem
surfaces between the loaded and unloaded surfaces (Figure N.2A).
7) The cross sections through the isosurfaces along the x, y, z planes could also be
compared, to determine the change in surface deviation through each section. On
the ‘Home’ tab, click ‘Section Through Object’ to select the plane along which
analysis is desired. For comparison of stem deviation along the length of the
stem-cement interface, the Y-Z plane is chosen.
236
8) To analyze the sections through the object, select ‘2D Dimensions’ from the
‘Home’ tab at the top of the program window. This tool allows the user to select
various locations along the reference (unloaded) and test (loaded) profiles, to
determine the vertical and horizontal deviations of the test (loaded) surface
relative to the reference (unloaded) surface (Figure N.2 B).
237
Figure N.2: Geomagic®
Surface Deviation Analysis
(A) The 3D Compare color plot shows the deviation in the stem and bead surfaces
between the unloaded and loaded condition, and (B) the 2D Dimensions from the section
through the Y-Z plane of the cemented stem and beads measure the vertical and horizontal
translation of the surfaces. The red profile represents the reference (unloaded) surface,
and the black profile represents the test (loaded) surface.
238
Appendix O: Uncertainty of Stem and Bead Displacements
This appendix describes the scan-rescan technique used to determine the uncertainty in
the measured displacement of the cemented stems and beads embedded within the
cement, described in Chapter 7.
Introduction:
The CT imaging methodology described in Chapter 7 of this thesis provided a
useful technique for analysing motion along the stem-cement interface. However, the
apparatus used for testing (i.e., holding jig, aluminum tube, titanium stem, steel beads,
aluminum load cell) incorporated metal components, which were carefully chosen to
meet CT-compatible requirements, but may have resulted in some error from inherent
metal artifact. In particular, the embedment of high density steel beads used for analysis
of cement motion in Chapter 7 (Appendix J) could have imposed artifact errors in the CT
generated surface profiles, which would result in uncertainty in the cement motion
measurements. As such, it was necessary to determine the amount of error imposed on
the stem and bead displacement measurements as a result of metal artifact contribution.
Materials and Methods:
Three smooth implant stems, similar to those used in Chapter 7, were potted into
square aluminum tubes using a 0.2% mixture of orthopaedic bone cement and steel beads
(Appendix J). The stems were left 24 hours to cure, and subsequently used for scan-
rescan testing.
The stems were secured within the custom built loading device (in the same
configuration expected for testing, without the application of a load) and two consecutive
scans similar to those described in Chapter 7 (x-ray potential of 120kVp; 40mA tube
current; duration of 8 seconds; image volume was 1024 mm x 1024 mm x 360 mm) were
acquired. No changes were made to the set up between scans.
239
The two scans for each of the stems were analyzed using the methodology
described in Appendix M. For Geomagic® analysis, the first scan was set as the
reference, with the second scan (i.e., re-scan) was set as the test. Scan and re-scan
surfaces for the three stems were compared to determine the average deviations of the
stem and bead surface between the scan and re-scan surface profiles.
Results:
Comparison of surface deviation among the three cemented construct showed that
on average, the stems (i.e., head, uncemented shaft and cemented shaft) showed an
average surface deviation of 0.014 ± 0.010 mm, and beads (Region 1 and 2, above and
below stem) showed an average surface deviation of 0.018 ± 0.011 mm (Table N.1).
Discussion and Conclusion:
Based on results from this pilot test, it was determined that uncertainty in the stem
and bead displacement measurements using CT-scanning and Geomagic analysis was on
the order of tens of microns. However, from this pilot study, stringent criteria for
analysing surface deviation of the stem and bead surfaces were developed, to exclude any
inconsistencies in surface morphology from the motion measurements. This included
analyzing contralateral edges of the stem and bead surfaces (i.e., top and bottom of stem
2D surface profile, top and bottom/ left and right of bead 2D surface profile), and finding
the average surface deviation value to represent motion. If contralateral deviations were
not similar, the measurement was excluded from the analysis.
240
Table O.1: Deviation in Stem and Bead Surfaces between Scan- Re-scan Tests
Stem #
Stem
Head
(mm)
Un-
cemented
Shaft
Cemented
Shaft
(mm)
Cemented
Shaft
(mm)
# 1 # 2 Region 1 Region 2 Region 1 Region 2
1 0.013 0.008 0.006 0.010 0.010 - 0.008 0.008
2 0.009 0.008 0.008 0.027 - 0.019 0.025 0.021
3 0.019 0.011 0.008 0.040 - 0.029 0.025 -
Average 0.013 0.009 0.007 0.025 0.010 0.024 0.019 0.014
SD 0.005 0.002 0.001 0.015 - 0.007 0.010 0.009
Bead Motion
(Above Stem)
(mm)
Bead Motion
(Below Stem)
(mm)
241
CURRICULUM VITAE
NAME: Yara Kareen Hosein, B.Sc.
EDUCATION:
2008–2013 Doctor of Philosophy Candidate, Biomedical Engineering Graduate
Program, The University of Western Ontario, London, ON
2004–2008 Bachelor of Science, Honours Specialization, Medical Biophysics,
The University of Western Ontario, London, ON
HONOURS AND AWARDS:
2010–2014 CIHR Graduate Fellow in Musculoskeletal Health Research, Joint
Motion Program (JuMP)—A CIHR Training Program in
Musculoskeletal Health Research and Leadership
Fall 2010 Graduate Student Teaching Award, Department of Mechanical and
Materials Engineering, The University of Western Ontario,
London, ON
2008–2012 Graduate Thesis Research Awards Fund, Western Internal Grants
Competitions, The University of Western Ontario, London, ON
2008–2012 Western Graduate Research Scholarship (WGRS) for outstanding
academic achievement, The University of Western Ontario,
London, ON
2008 Gold Medalist (highest academic standing), Honours Specialization
in Medical Biophysics (Physical Science Concentration), The
University of Western Ontario, London, ON
2007 A.K. Knill Residence Staff Leadership Award, Perth Hall, The
University of Western Ontario, London, ON
2006–2008 Dean’s Honour List, Faculty of Science, The University of Western
Ontario, London, ON
2006 Gold Medal, Excellence in Leadership Award (volunteering and
community service), The University of Western Ontario, London,
ON
242
RESEARCH EXPERIENCE:
2008–2013 PhD Candidate; Graduate Research Assistant, Biomedical
Engineering Graduate Program, The University of Western
Ontario, London, ON
2011–2012 Interdisciplinary Research Collaborator, Graduate Department of
Orthodontics, Schulich School of Medicine and Dentistry The
University of Western Ontario, London, ON
2007–2008 4th
Year Honours Medical Biophysics Thesis, Imaging Research
Laboratories, Robarts Research Institute, The University of
Western Ontario, London, ON
May–Aug 2007 Student Research Assistant, Imaging Research Laboratories,
Robarts Research Institute, The University of Western Ontario,
London, ON
Jan–Apr 2007 Student Research Project, Jack McBain Biomechanical Testing
Laboratory, The University of Western Ontario, London ON
TEACHING EXPERIENCE:
Winter 2012; Graduate Teaching Assistant, Finite Element Analysis,
Winter 2011; Department of Mechanical and Materials Engineering
Winter 2009 The University of Western Ontario, London, ON
Supervisor: Dr. Paul Kurowski
Fall 2011; Graduate Teaching Assistant, Product Design & Development,
Fall 2010; Department of Mechanical and Materials Engineering
Fall 2009 The University of Western Ontario, London, ON
Supervisor: Dr. Paul Kurowski
Winter 2010 Graduate Teaching Assistant, Introductory Engineering Design &
Innovation Studio
Department of Mechanical and Materials Engineering
The University of Western Ontario, London, ON
Supervisor: Dr. Paul Kurowski
243
SUPERVISORY EXPERIENCE:
May–Aug 2010 Meghan Clynick: First Year Undergraduate Student,
The University of Western Ontario (NSERC, USRA)
PUBLICATIONS IN REFEREED JOURNALS:
1. Y.K. Hosein, G.J.W. King, C.E. Dunning. “The Effect of Stem Surface Treatment
and Material on Pistoning observed in Ulnar Components of Linked Cemented Elbow
Prostheses” Journal of Shoulder and Elbow Surgery. 2013 [E-pub ahead of print;
PMID: 23668920]
2. Y.K. Hosein, G.J.W. King, C.E. Dunning. “The Effect of Circumferential Grooves
on the Stability of Cemented Joint Replacement Systems” Journal of Medical
Devices. 2013 [In Revision: Paper # MED-13-1022]
3. Y.K. Hosein, G.J.W. King, C.E. Dunning. “The Effect of Stem Material and Surface
Treatment on the Torsional Stability at the Metal-Cement Interface of Cemented Joint
Replacement Systems in the Upper Limb” Journal of Biomedical Materials Research:
Part B Applied Biomaterials. 2013 [Under Review: Paper # JBMR-B-13-0196]
4. S.R. Fetterly, Y.K. Hosein, C.E. Dunning. “Development and Validation of a Strain
Gauge Embedment Methodolgy for Use with PMMA Bone Cement” Journal of
Biomechanical Engineering. 2013 [Under Review: Paper # BIO-13-1207]
5. Y.K. Hosein, M.P. Clynick, S.D. McLachlin, G.J.W. King, C.E. Dunning. “The
Effect of Stem Curvature on Torsional Stability of a Generalized Cemented Joint
Replacement System” Journal of Applied Biomaterials and Functional Materials.
2012 [E-pub ahead of print; PMID: 22798236]
SCIENTIFIC POSTER PRESENTATIONS:
1. Y.K. Hosein, G.J.W King, C.E. Dunning. “The Effect of Intramedullary Stem Surface
Finish on the Torsional Stability of Cemented Joint Replacement Systems”. Canadian
Orthopedic Research Society Annual Meeting, Ottawa ON, Canada 2012.
244
2. A.L. Smith, Y.K. Hosein, A. Tassi, C.E. Dunning, “An In-Vitro Study Evaluating the
Fracture Resistance and Insertion Torque of Self-Drilling Mini-Implants upon
Insertion into Synthetic High Density Mandibular Bone”. Canadian Association of
Orthodontists 64th
Annual Scientific Session, Ottawa ON, Canada 2012.
3. Y.K. Hosein, M.P. Clynick, S.E. Takaki, S.D. McLachlin, C.E. Dunning. “The Effect
of Intramedullary Stem Curvature on the Torsional Stability of Cemented Joint
Replacement Systems”. The Dr. Sandy Kirkley Centre for Musculoskeletal Research -
Poster Competition. London, ON., Canada 2011.
4. Y.K. Hosein, M.P. Clynick, S.E. Takaki, S.D. McLachlin, C.E. Dunning. “The Effect
of Intramedullary Stem Curvature on the Torsional Stability of Cemented Joint
Replacement Systems”. CIHR Joint Motion Training Program in Musculoskeletal
Health Research and Leadership: Annual Retreat University of Western Ontario.
London, ON., Canada 2011.
5. Y.K. Hosein, M.P. Clynick, S.E. Takaki, S.D. McLachlin, C.E. Dunning. “The Effect
of Intramedullary Stem Curvature on the Torsional Stability of Cemented Joint
Replacement Systems”. Summer Bioengineering Conference, American Society of
Mechanical Engineers (ASME), Farmington, PA., USA 2011.
6. Y.K. Hosein, M.P. Clynick, S.E. Takaki, S.D. McLachlin, C.E. Dunning. “The Effect
of Intramedullary Stem Curvature on the Torsional Stability of Cemented Joint
Replacement Systems”. Orthopedic Research Society 2011 Annual Meeting, Long
Beach, CA., USA 2011.
7. Y.K. Hosein, S.D. McLachlin, G.J.W. King, C.E. Dunning. “Development of
Methodology to assess stem surface finish in cemented implant loosening”. Summer
Bioengineering Conference, American Society of Mechanical Engineers (ASME),
Naples, FL., USA 2010.
8. A. Wheatley, S. McKay, L. Mathew, Y. Hosein, G. Santyr, D.G. McCormack, C.
Licskai and G. Parraga. ‘Hyperpolarized 3He Magnetic Resonance Imaging of
Asthma: Short-term reproducibility’. London Imaging Discovery, London, ON.,
Canada 2008.
9. A. Wheatley, S. McKay, L. Mathew, Y. Hosein, G. Santyr, C. Licskai, D.G.
McCormack and G. Parraga. ‘Hyperpolarized 3He Magnetic Resonance Imaging of
Asthma: Short-term reproducibility’. Robarts Research Day, London, ON., Canada
2008.
10. A. Wheatley, S. McKay, L. Mathew, Y. Hosein, G. Santyr, D.G. McCormack, C.
Licskai and G. Parraga. ‘Hyperpolarized 3He Magnetic Resonance Imaging of
Asthma: Short-term reproducibility’. Society for Photonics and Instrumentation
Engineers (SPIE), San Diego, CA., USA 2008.
245
INVITED PODIUM PRESENTATIONS:
1. Y.K. Hosein, G.J.W King, C.E. Dunning. “The Effect of Stem Surface Treatment and
Material on Pistoning observed in Ulnar Components of Linked Cemented Elbow
Prostheses” Ontario Biomechanics Conference, Barrie, ON, Canada 2013
2. Y.K. Hosein, G.J.W King, C.E. Dunning. “The Effect of Intramedullary Stem
Surface Finish on the Torsional Stability of Cemented Joint Replacement Systems”.
Ontario Biomechanics Conference. Barrie, ON, Canada 2012
SCIENTIFIC MEMBERSHIPS:
2011–2013 Biomedical Engineering Society (BMES)
Student Member
2012–2013 American Society of Mechanical Engineers (ASME)
Bioengineering Division (BED)
Student Member
UNIVERSITY SERVICE AND VOLUNTEER WORK:
2010–2013 Outreach Coordinator, CIHR Joint Motion Program (JuMP), The
University of Western Ontario, London, ON
2008–2013 Biomedical Engineering (BME) Student Committee Member, The
University of Western Ontario, London, ON
2011–2012 Graduate Student Representative, Biomedical Engineering (BME)
Undergraduate Task Force, The University of Western Ontario,
London, ON
2009–2011 President, Biomedical Engineering (BME) Student Committee,
The University of Western Ontario, London, ON
2008–2009 Volunteer with Western Serves, The University of Western
Ontario, London, ON